I sat front row for David's talk yesterday morning. Having talked with him the day before, I’ll admit… I wasn’t particularly surprised (nor unprepared).
A bit of a field report from #RailsWorld: the vibe here is far from doom and gloom. Quite the contrary. Wherever our industry is headed, most of us are still employed as menders… tending to systems that customers rely on and businesses are quite happy to keep paying for.
Most of us aren’t waking up to a blank canvas and designing the architecture of the future. We’re inheriting decisions made years ago, updating old patterns, working around constraints, and keeping this shit running reliably.
I think these newer tools give us an opportunity to wonder a little more about the systems we’ve inherited. To unpack why things work the way they do. To tinker with assumptions we haven’t had the time, confidence, or permission to revisit… and share what we learn so the next person, or agent, has an easier time.
That deployment model we picked eight years ago? Worth another look. Some of our web apps probably wish they were native apps. And there are plenty of architectural decisions we’ve been living with mostly because… well, we’ve been busy living with them.
There’s plenty of understandable anxiety about what these tools mean for our work. I’m increasingly curious about what they give us permission to revisit.
Once my keynote is published, I’ll share more about a little programming-language-adjacent framework I’ve been working on for approaching exactly this kind of curiosity.
There is a world of difference of having tens of millions of dollars and paying hundreds of dollars a month for AI subscriptions vs someone making 100k / year doing the same.
For the DHH's of the world that's not even something to think about, but for most people who are working (even in tech), now we're talking 2-8%+ of your income depending how much you use AI. Just to have the "luxury" of being able to experiment with it. I use air quotes on luxury because I personally find a lot of AI results to be quite bad, but then again, I'm not spending hundreds of dollars a month on the premium models.
I make 55k working at a dental office and a restaurant and I'm paying 200$ a month on open source models because I'm having an absolute blast with them.
Plenty of jobs have expenses built into them. Maybe a subscription is more than it used to be, I had great jobs on a used 200$ thinkpad, but it's still cheaper than being a photographer. And 100k is more than most jobs.
Software dev, even if it's all agents now, is still pretty dang economical
Have you tried OpenRouter and the DeepSeek-like models? If you try to optimize a little you could probably slash those costs.
I send a ton of requests and even reaching $10 for the month would take more calls than I have capacity to review and use. And as I said, that's for a ton of requests.
20 a month and using sonnet mostly for random fun stuff.
At work, it seems a bit different - I'm certainly having to declaudify things it can be hard starting there's a lot to dig into - you can see it's overwritten but need to understand every detail to sort it out.
It's easier to enjoy it if you don't see how the sausage is made.
I am definitely not having a blast at work with LLMs. Everyone in the team has become simple rubber-stampers for whatever the AI puts out. There are mountains of documentation which gives a false sense of understanding of how things work, but the code repository itself could have been written in Brainfuck and it would make almost no difference. And we keep getting told by management that this is fine. We've become simple janitors in the slop hall.
Everyone in the team has become simple
rubber-stampers for whatever the AI puts out.
Whoever figures out better ways to:
1. Present a big AI PR to a fellow human for review
2. Understand a big AI PR, as a human reviewer
...is going to be a hero. This is one of the biggest unsolved problems of this new era.
I've been experimenting with this, but it's early days. My efforts are mostly along the lines of having AI generate artifacts that guide the reviewer through the PR in execution flow order, explaining the reason for changes along the way.
I don't think that's the way forward. We don't accept mega-PRs from people, why accept them from machines? The best practice of small, focused commits - whether it be from a person or a LLM - will win out in time once we're past this manic phase.
You think 2-8% is a problem? There's a large world outside the US. "Hundreds of dollars" per month is most of my salary, I can afford that only by going hungry half the time.
Probably going to if I want to keep my job, since I'll be making about the same amount of money moving boxes or cleaning toilets, only in much less comfortable conditions.
It's essentially trying to turn programming into a micro-transaction based pay to play / pay to win outcome, a concept that was gaming related for decades.
You can pay $5 for deepseek 4.1 flash and be on the pareto frontier. And you can pay between $20-200/month for a huge amount of usage of the absolute frontier models. I was concerned about costs a year ago but with the proliferation of great cheap models, I think it's no longer a concern at all. I'd be fine if I could only use 4.1 flash.
I think the fear, threat, and valuation, of the frontier labs is that they will cross some line and have models that are infinitely smarter and more capable than everyone else. If that should happen, your deepseek account would be meaningless.
But I also think that premise is bullshit and not gonna happen.
The premium models are worth 10% of your income at least.
There is no future without these models. They truly are an entirely different league than the cheap stuff. You can move mountains with Astra or Opus 5.5.
10% of a software engineer salary is a lot of money. That's just ridiculous.
Like the other commentor said, its just going to be a net negative, because whatever productivity increase you supposedly do get, you will see exactly 0 of that translated into salary increase.
Have you tried to negotiate a 10% pay raise because “you can move mountains with Astra or Opus 5.5”? I’d wager you’re more likely to get a “the fuck do we need you for then” than that pay raise.
The whole AI issue is and always has been a social question. It's hardly a ridiculous concept to have a skilled worker control a $10k budget to buy tools. In many fields that is normal.
Previously, the tools in our field were free or trivial cost. Seems like that has changed, that's OK, there are models to solve that.
I teach a couple courses currently, there is an enormous gap between what my students can create with AI and what I can create with AI. I don't think it's that much smaller than the difference of what we can create without, the only difference is that the schedules are compressed.
I swap between Fable and GLM 5.X and both work fine. Fable tends to do stuff that I didn't want it to do, and claude is probably a bit more capable as a harness (though I haven't spent the effort configuring pi properly yet so that's probably on me). Otherwise it's hard to spot the difference.
How can Rails live long if no one creates new things with it? Our web applications aren't like the Pyramids at Giza. DHH literally said that even without a blank canvas (Hey's backend) they were still able to re-write the backend in Rust and are not looking back.
That sure sounds like any savvy business should pivot away from Rails and Ruby in general as fast as they can, assuming they have tokens to burn. Personally, I built my career on Python, but any new thing I build is almost always in Go or Rust. Why mess with an interpreter if you don't have to?
Having built two apps in the last two years using Rails (one making front page on Hacker News and has 10,000 plus users), there is a sweet spot that Rails hits that other frameworks don’t.
The fact that you get so much out of the box that just works and you don’t have to do much beyond configure it reduces a ton of headaches.
Absolutely would rather be using Go instead of Ruby but for apps of a certain kind and size, the omakase approach of Rails still wins out.
He clarified today, saying he would build any new app today in Rails. Should that app become so popular that it would benefit from Rust, he would rewrite it in Rust.
According to the schedule, today had a keynote on "AI and the future of Ruby & Rails — a chat with Matz & DHH". I suspect that might have clarified, but sadly they don't seem to have released a video (yet?)
I agree with him. All software should be written in safe languages, but there is a wide selection available. For new projects you should pick a language that does as much for you as possible. Ruby and Rails both go out of their way to provide the developer with as much power as they can. They make development less expensive. That’s exactly what you want when you are first building something, before you know exactly what you will need.
Later after you have lots of users and your hosting costs are rising, then you can rewrite it in a language that gets you closer to the machine. You won’t need to do any exploration at that point, and you’ll already know where the performance hot spots are. Of course I recommend Rust for that, but that is merely my personal preference.
You can accuse DHH of many things, but not of grifting, insincerity, political politeness/correctness just not to offend people. The guy speaks his mind and is unapologetic about it.
The drastic change in the ethnic/cultural makeup of people leaving in many Western countries is a fact. In Finland it is +70% foregners only in the past 10 years. Do you want to dispute stats? Write to the Finnish department of statistics.
Or are you arguing that this is a natural process and politicians can't influence it? Wrong. Pre-post Brexit stats show the opposite. As do the stats from Denmark, after politicians started working on the issue.
You are fine with stats, but don't like DHH's opinions about those facts? Well, those are opinions and everyone has a right to them. Opinions and feelings based on morals can't really be proved incorrect.
Or are you tone policing?
So what exactly is the actual problem? "Someone on the internet has wrong opinions and is not a carbon copy of me, politically"? What a silly and totalitarian behavior to call other people dirty names!
Stop calling anyone with differing opinions a racist or xenophobe. It really feels like these activists are deliberately misusing powerful words in order to weaken them (if everyone is called a racist, no matter how undeservedly, then no one is). Or they are simply stupid and do not understand the reaction they are causing with their totalitarian tactics.
> Why mess with an interpreter if you don't have to?
I mean I haven’t touched Ruby in years but if you’ve got tokens to burn, who cares? Compute is pretty cheap these days and Ruby is usually fast enough for web stuff, and as far as I can tell 2nd hand the robots write it just fine.
And I say this as someone building a strongly typed programming language that compiles to native code.
I actually have no problem with him even if he directly say it out loud announcing the death of Ruby and Rails. And I am not even against agents taking over some jobs meaning less jobs for current developers. That is part of how new technology has always been, replacing some and we move on.
It is just that it shouldn't be done in the opening keynote of a conference. I thought it was very irresponsible to the other smaller sponsors and those who paid to join. I know Shopify fund most of it and is likely on the same stand.
> In the meantime… keep showing up. Keep wondering.
> Long live Ruby. Long live Rails.
> p(bloom)
These aphorisms try to put a more positive spin on the keynote, but they miss the point that made it so disappointing for a lot of people.
They travelled to a Ruby on Rails conference because they wanted to hear RoR talks. The keynote was all about AI and Rust and ended with “don’t be losers” about the AI debate.
Ruby on Rails content not found.
I suppose the die hard fans who have been around this long and who continue to travel to see DHH speak have aligned themselves with him as a personality figure, maybe more so than with Rails itself. Outside of that bubble, it just feels sad that one person used the conference to do yet another controversial move in an inflammatory way while ignoring the topic everyone travelled for. I say this as someone who uses a lot of AI and also agrees that programming work has changed. It’s just so insulting to be surprise-lectured about it with condescending “don’t be losers” comments when the topic was supposed to be Rails.
There was an opportunity to talk about what AI meant for the future of Rails, but instead he went off bragging about rewriting Hey in Rust. Rust has been my primary language for years, but I don’t want to be lectured about it in a Rails keynote while Rails discussion is nowhere to be found.
I’ve made a similar point to my colleagues recently about cross-platform vs native apps. If we’re using AI anyway, do the arguments we made years ago still hold today?
I know some people might disagree, mainly because AI isn’t a new architecture or a new programming language in the traditional sense. But I still think it’s worth revisiting the assumptions and having the discussion.
Put aside the usual web-dev perspective for a moment. As developers, we build for our users.
The web and native apps serve different purposes. Users do not use a website the same way they use a mobile app, and the trade-offs are different.
I’m getting tired of web developers failing to acknowledge that and treating web technology like a hammer that needs to be applied to every nail they come across.
And especially when people say “the web”, what they often really mean is Chrome and the Chromium ecosystem. A lot of web apps barely get tested properly on Firefox or Safari, which only reinforces Google’s monopoly over the web platform.
Yes, app stores have their problems. That does not make the web the right answer for every product.
For a lot of companies it is not a factor because they have free app complementary to their services.
Main point is that even if you could slop out some code to have a mobile app there is enough of additional burden that makes it not really worth it when you can get away with web version that is good enough and you don’t have to care about all that bs coming with publishing and maintaining a mobile app or to be specific you need to maintain exactly 2 apps.
Depends a lot on the token budget I think, different situations for established businesses or strapped founders vs side hustles or hobbyists. So in a way you outgrow Rails or frameworks like it once you can afford a software factory on autopilot 24/7.
That worldview assumes that technology is like rain or an asteroid shower, and we just have to accept it. Its not. Humans make technology and decide what is going to be used.
No technology is inevitable. Technology isn't nature.
The constant has always been tools that maximize developer productivity. What's changed is that LLMs are now the best way for developers to tell machines what to do. We've all been given a much bigger lever we can use to spend our waking hours (and even sleeping hours if you run agents overnight) to move even bigger mountains.
The market being what it is, I took a job in ColdFusion last year.
There are people here who have worked in the same CF codebase for over twenty years. They might start and finish their careers in a language everyone else forgot about decades ago.
I think Ruby has enough momentum that career Rubyists will have a home for as long as they want one.
The foundations yes, the rest only if you never change between RDMS products.
On my 50s, learnt SQL around 1996, have used SQL across Access, Informix, DB2, Sybase SQL Server, MS SQL Server, Oracle, MySQL, Postgres, PrestoDB, Neo4j.
They’ll be getting nervous too at some point the agentic port will come for cobol. The past hesitation and risk aversion will subside once the less risky stuff is addressed. Once someone big enough sets a precedent, it will gain momentum
To be fair, this is true of most of AFK world. Things start seeing a lot less doomy and gloomy when you're actually outside talking with other humans. Seems to mostly be online and various darker corners where they're dooming and glooming.
I get why DHH would be really happy about getting rid of engineers. It's less clear how this isn't a little bit gloomy for the engineers.
Is the idea that they'll switch to doing "product owner" roles overseeing agents? I don't see that employing the same number of people or paying as well as engineering used to.
To be honest I expect that that role to suit different people to those that used to be engineers.
If this feels gloomy to some people, it's because it is.
I don't think that hiring the engineers was ever an option. He said that he couldnt maintain six native apps before agentic coding. The business never even considered expanding headcount by 6X to do it, it was a complete nonstarter.
There is no fixed pie of work that AI is taking from us. In this circumstance, increased productivity caused the company to just make more stuff, not lay off people (or avoid planned hires).
To clarify, my point isn't that the market for software won't expand due to the greater productivity unlocked by AI.
It's that the benefits of that productivity won't go to the kinds of people that previously wanted to stay in engineering their whole career.
I'm not sure the Jevons Paradox applies because you're not just making engineering more efficient you're replacing it altogether with something much more commoditised.
People will gain from this but it won't be the same people.
Likewise, consumers will eventually reduce the amount they’re willing to pay for software so it will balance out. It will be a blip in time we look back on. Like, remember when elevators used to employ an operator. Now we just push a button and happily do it without another person’s labor.
Well at least in this particular case, it was the same engineers on the team. He specifically said he tried getting UX people to vibe code things and the results didnt work
Jevon's paradox has to do with consuming resources, not employment. There are plenty of examples of fields getting more efficient and employment falling. Farming and manufacturing being 2 examples.
You're conflating different categories of activities. Farming and manufacturing are amenable to automation precisely because not much real thinking and communication is required.
It's an interesting example in that a switch from web/React -> native seems like Jevons paradox from the perspective a developer considering "amount of effort I'd need to put in"
But Jevon's paradox is about consumption of a resource increasing. The resource being "developer effort" but the new supply all running inside GPUs and dev machines. He actually says he decreased their server deployment by 90%.
So on net, did they basically deliver their users a much nicer but functionally equivalent app, which in total now consumes less resources / spends less $ into the economy? And then the $10T question is to what degree this is representative of AI's effect.
the algorithm is programmed to deliver it. Better engagement metrics with misery and anger than joy and hopefulness. (yes I know that you will reply 'but I prefer the joy and hope', but that's not the case across the wider population)
And I think the online reality is much closer to the truth than AFK reality. I can't think of a single founder I've talked to in the past two years that's building with rails. That doesn't mean rails dies overnight, but it does mean it's simply all downhill from here.
Promoting his rails services where the creator and biggest defender of rails replaced rails with an AI generated rust version without writing a single line of code.
His comment is just damage control and biased, it did not state he's heavily invested financially in rails. Not everyone knows who he is, I didn't.
Concerns with his politics aside, I do think there's a truth to what he's talking about here, and that he is just spelling out the reality that developers are, or shortly will be, facing. For many this will be deeply uncomfortable to hear.
It is notable that his perspective in this talk is very much from a developer-user side rather than someone who is responsible for the framework itself. That surprises me, and I suspect it is not a good omen for Rails.
For all of this embrace of agentic development, there is nothing here on how they are adapting the framework for this new agentic development reality. Agents do currently work well with Rails, but there's nothing here pushing things forward as best as I can see.
I know for my own projects I've largely moved to Elixir/Phoenix, for similar reasons to his use of Rust... I didn't want to have to learn it, but now I don't have to and I get to benefit from its strengths.
I think a lot of people haven't actually watched his talk and don't understand what he is saying. If you watch the talk, he is not anti-rails. He is just very pro- agentic coding.
What he was saying in his talk was that his customers (people using Hey for email and calendar) never really wanted hybrid web-native apps, they want native apps (like most people prefer on mobile). His business could never afford to have a dev team that built and supported six different native apps, so they always built using hybrid web-native stuff using Turbo Native.
Now that his development team has access to agentic coding, they can build and maintain six mobile native apps. For his app, the back end part then becomes small and simple and they can reimplement it in rust to reduce server cost.
His point wasn't that rust is better than ruby, or anything like that. His point was that agentic coding allows smaller teams to do bigger things. For most people building on rails, their customers don't need native mobile apps (I know mine don't) and for them Rails is a fantastic choice for agentic coding.
Rails is so into agentic coding that they publish benchmarks of how effective different models are for building with rails:
TLDR: DHH isn't anti-rails, he is pro-agentic coding. Rails is great with agentic coding. AI doesn't mean layoffs, it means your teams can do much more than they could do before.
It seems to me, so far, everyone moved one level up.
- A layperson can now build one a solo developer can do
- A solo dev can now build what an agency of web development can do
- An agency of web development can now build native apps similar to what large corporation can do
- Large corporation can now tackle even more ambitious projects
Which makes sense given what got automated. This is with the current tech, people still exploring the full potential, but if that turns out to be the case then I'd argue the software market will grow substantially and we are just having an adjustment period.
This matches my experience. My solo projects are now bigger than I ever would have accepted alone. The one I’m going to finish soon is one I actually refused a year ago because it was too big.
It’s the top of the chain that I wonder about. Agentic development is so fast (and in the process I’ve developed so good) that I’m thinking there is actually a maximum size a company doing this work could get. Beyond that top end size the management overhead is what’s limiting or the market size total. Maybe not, maybe mega dev houses will find other problems to solve, but I have a hard time imagining it.
I do suspect there will be many mega projects. Things like new operating systems, new apps ecosystem (not just apple/android), new virtual reality environments, massive interconnected web of agents, in-house models etc, also alot of re-writes and update of existing infrastructure.. competitors to existing social networks and many established players..I think the entire software ecosystem will grow larger, wider and deeper from all directions.
From what I have seen the pie will get way larger, but there is growth pain and adjustment period for sure.
But to your point, I also think solo devs/freelancer are now seeing the most gains, that's because they can adapt the fastest and they don't need to deal with existing overheads.
Larger organizations might need more tooling/agents and better internal processes before they can see the full potential. I don't think ww, as an industry, are there yet, this tech started truly working less than a year ago.
It looks like he spent about 4 minutes out of the 1:03 making that point.
I can't but help think that he would have spent longer on it if that's what he really thought.
I don't think he believes in Rails any more.
It's clear that he was more excited about thing that are not ruby or rails and now considers rust better suited to the output of AI.
To be honest, I could see rust taking over from typescript/javascript as an output language for the web too. If no one needs to read source code any more why not use rust for everything?
I don't think anyone should really believe in Rails at this point. It was a monumental leap forward at the time but other stacks have shown that there are better ways.
>It's interesting how this stuff gets spun on social media by people who don't watch the content.
I watched it live, and a few actually in the show have very different perspective to what you suggest, they were just keep it to themselves and not posting it on twitter. Some did, but then deleted it.
And no one is questioning he doesn't like ruby anymore. But it was obvious there was nothing much said about the future of Ruby and Rails.
When I started as rails dev in 2013, it was really the top solution for producing MVP fast. Now, I don’t think that’s the case and most of the incentives are gone if we are no longer writing code by hand
So what would you pick for producing an MVP web app fast that you also want to run in production, maintain, patch and scale that has a solid eco-system of pre-built libraries you can just plug in to solve problems?
I strongly prefer Rails with agentic development. Rails comes with nearly everything you need to make a full-featured application, and that makes it easier and faster to use with agents.
No other framework comes close, and telling me that I can have my agents cobble together this same baseline, badly, from a hodgepodge of new code and random libraries is not compelling.
At least within my circle, that’s not really the case. Instead, what I’m seeing is more of a hiring freeze, with the budget being used to empower existing employees through AI. Which I think is a more reasonable take.
That's often layoffs in disguise. If you want to cut 10%, but you don't want to do it as a single layoff, because that's expensive, can trigger certain laws depending where you are, and might disrupt teams, who do you cut?
Instead you can do layoff by attrition, you stop hiring, and you don't backfill seats, as people naturally leave or get pip-ed, the team shrink, you don't rehire into the position, and in a year or two, you cut down by 10%.
Also, I wouldn't trivialize a freeze in hiring growth, it means to get a job you have to wait for someone to retire, and also wait for people looking for jobs with more experience as they're more likely to fill the roll to also retire. Basically there's now way more demand for jobs than hiring, so that will drive compensation down for those currently employed.
Adding on to your point about attrition: if internal promotions in engineering are decided by your ability to be a force multiplier, to facilitate cross team collaboration, mentor and help others to skill up, how do any currently employed engineers demonstrate that with no one new to mentor or train? How do you collaborate when teammates work with AI tools more and coworkers less? Would they just have an easier time getting into a higher responsibility role looking elsewhere?
> We can't hire or give anyone raises, but we will expect you to be more productive via throwing money into the slop void.
i see your point and i want to tell you it's pretty much the same with cloud computing, which is so much more expensive than running your own hardware... and yet nobody's really questioning cloud computing (except the oxide people, but they're still tiny)
the company i work for could probably spent half its yearly cloud bill in hardware and be fine for 3-5 years.
guess who discovered the same thing and booted google cloud off their books? dhh
as a solo: you can build and ship more in your own time
as a startup: you can iterate and scale more with more velocity and less growing pains
as an enterprise: people with ideas but not a technical background can put their tokens where their mouth is and build their own interfaces, personal apps, ask their own questions, ship their own features without having to navigate layer 8 and play games for resources
Why does it sound like layoffs? If you keep the team size the same and use the productivity increase to deliver more business value, how is that layoffs?
Thus layoffs, no need to keep as many employees around.
There is no MBA like magical growth, the amount of customers is a physical limitation of inhabitants on planet Earth that are willing to actually use a specific piece of software, regardless of how productive it was made in first place.
If I figure out a way to make my workers produce 10 times more widgets each, it doesn't automatically follow that there is a market for the extra 9 widgets. Otherwise I would have hired 9 extra workers before that invention to fulfill the market demand.
Making the widget cheaper doesn't necessarily create demand either. I will not buy 10 times more toothbrushes or writing pens if their prices suddenly decreased to 1 cent.
The demand for software or features will not increase at the speed of our ability to produce them.
That is the fallacy of all those MBA degrees with exponential growth expectations, they never take into account the physical limitations of goods, like your very good examples.
If every engineer is 10-100x more productive with AI, and one company slashes its workforce to maintain the same level of productivity at reduced cost, it will lose out to its competitor who retains the same workforce and does far more with it.
Right now, the vibe at my company is not “we can save so much money with AI” it’s “here’s an infinite budget, use all the tokens you can to expand our moat because we’ve never been more sensitive to competitive disruption than we are now”
One of the things I admire most about Basecamp is that they are a bootstrapped, profitable business. There's no shame in running a business this way. The world would be more resilient if more business owners grew sustainably instead of soaking up endless invester funding and losing money year after year before enshittifying in a desperate attempt to repay their overlords.
Here, I will give you my upvote! That's how Issuetrak is. The founder (me), still owns a majority of the stock, there's no outside investment, and we have customers who've been with us more than 20 years. Multiple employees as well!
>His business could never afford to have a dev team that built and supported six different native apps
I have a hard time believing that, considering his hobby of racing cars and commissioning one-off hypercars. Unless it's that his business could never afford that dev team after DHH takes his salary.
It’s not his message, it’s the forum. Why host a rails conference, that he explicitly created because he was silenced from the main one…to them not talk about rails?
Why not adapt rails for this next generation?
Why do it this way? It’s like walking into a temple to yell about Jesus. It was absurdly obnoxious.
> TLDR: DHH isn't anti-rails, he is pro-agentic coding.
That's an opinion, not a summary though. The way how DHH behaved is really awkward in the last years. See the "let's shoot down wolves" blog entry here:
That DHH is into AI agents everywhere is easy to see, but if you go to rails-con, would you not expect this to be about rails as such, and people? Now it's about rust - and AI. Controlled by big corporations. Yikes - some people must be upset about having gone there to listen to DHH preaching about an agentic future.
Edit: I also watched parts of the recent youtube video about rails. I could not focus on the content because DHH's style was so strange. I've seen that style before - nothingburger talks. Look at phrases he used such as "I have retired from being a professional programmer/software engineer". This is all aimed to hype up "agentic coding is the future". Rails is dead - that video is all the proof one needs. DHH shifted gears to AI, and it is quite similar to when he first promoted rails. But now it is about AI, not rails.
Do the teams fully understand the code that they are committing? Are the teams reviewing each others code or are agents? When agents do all the coding, and perhaps the reviewing, and the patching what is left of the team? People lose connection to their work, and the social bonds between workers begins to erode.(1) DHH's speech seems like a weird eulogy for the entire Rails community, although many of us saw this coming several years ago.
Also, I hate native apps. FWIW most devs I know actively try to avoid them.
Most devs I know who hate native apps are either web devs trying to embed web tech into everything they touch, or they just hate Apple.
Most users don’t care. What they do care about is performance, and native is usually the easiest way to get there.
For your AI workflow questions, it depends. With or without AI, those decisions depend on the organisation’s direction, the team’s capabilities, and their preferences. Did your org have a single answer for the best programming language, framework, or workflow even before AI? There was never a single right answer to any of that.
I hate native apps. There are use cases better suited to native apps but often it's just not what I want.
You have little to no control over privacy, tracking or advertising compared to the web. That's left to the original developer and there's nothing you can do about it.
You can't do things like disable autoplaying videos. Often you can't even copy text without taking a screen shot. On a phone it's also much easier to zoom random parts of the screen in a browser compared to a native app.
There are legitimate reasons to prefer the web to native apps.
Users absolutely care, people hate downloading random shit on their phone they don't want but are forced to.
What has always been true is that users often have no choice or voice to express their actual desires. It's not like there is a public alternative here (when there should be).
> Rails is so into agentic coding that they publish benchmarks of how effective different models are for building with rails
To each their own, but I think using Rails in the era of agents is probably a mistake. It has 2 big problems IMO.
(a) A ton of footguns that an agent can and will miss on occasion, and as we manually review less and less code, we will also miss it, and
(b) security on the web has become increasingly important due to hostile AI agents, and while rails has great secure conventions (see (a)), it's still easier to screw up something and not catch it until after it's deployed because there are simply less guardrails in place (dynamic types, no compile, slow tests, generally bad performance profile).
Rails dev here, still using it, still plan on using it, use agents, don't see them making food gun decisions with rails apis just code.
Rails comes with the console by default, so production inspection of a bug is something the ai can do without a ton of bs to be set up, and that console can be forced to be auditable (using standard gems) or read only (using flags).
I used to love rust, still do sorta, but yeah I think this is really not a good argument against a tool that operationally is better at most things at being observed or poked or prodded or tested in production
Not sure if you’re speaking from experience or just being theoretical here. We’re using Rails and it’s been great. Maintenance and upgrades have been a breeze.
All this static vs dynamic stuff doesn’t really matter anymore. AI doesn’t care. You just need to close the feedback loop with tests, linting, a compiler, or whatever else gives it a clear signal.
Rust is used a lot. Elixir/Phoenix comes up often. You’re looking for something statically typed with string guarantees and as FP as possible to make unit testing easier.
Rust for sure. Builds become a problem eventually though. Elixir is much more interesting now that they added types (I think they're doing more with this still?). I quite like Gleam from a human (non-agent) perspective.
In my experience, Go. It was designed to be idiot-proof. It's very verbose, not very expressive, compiles fast for quick feedback loops, strict autoformatter, plenty of examples in models' training data.
the only substantive difference I've noticed is between languages with a proof system and those without, where with a proof system is orders of magnitude better for agentic work.
Same. I find that the combination of the BEAM runtime and the guardrails that you get with an Elixir project, functional style and particularly the consistent structure of using Ash with it for data, API and CLI lead to a very maintainable codebase even with a lot of AI.
Phoenix with LiveView bypassing the need for a separate frontend framework is also a huge win there.
> Concerns with his politics aside, I do think there's a truth to what he's talking about here
One thing the AI era has changed for me is how easily I spot Gell-Mann amnesia, of which this is an unfortunately strong example.
His politics, even putting aside their nature, are framed with deliberate misuse of facts and data; why would he be someone to listen to in any depth on any other topic?
Gell-Mann amnesia assumes a journalist is not an expert on _anything_. In this case, DHH is (according to some) an expert on software but it's much less clear he's an expert on politics. So I'm not sure Gell-Mann amnesia applies here, it's maybe more like Nobel disease, where you're an expert on one thing and assume it "transfers" to other fields.
> Gell-Mann amnesia assumes a journalist is not an expert on _anything_.
Incidentally, this is also the fundamental flaw with Crichton's original description of the effect which questioned the accuracy of reporting about international affairs. When it comes to politics, simply having access to important leaders and meetings and reporting on them accurately and fairly is the only "journalistic expertise" you need.
Of course, the even broader problem with the concept is that, in reality, there are some articles written by journalists that contain a reasonable understanding of the subject matter, and some articles that do not. So seeing an article written about your field of expertise that contains misunderstandings doesn't by itself tell you anything about the distribution of good and bad articles covering any particular field.
I have a background in sociology and urban planning. It's very clear that DHH is not an expert on politics. His blog posts demonstrate a very elementary understanding of the topics he covers. He doesn't even understand statistics.
People say "background" here, because the default assumption is HN readers have software expertise, but it's frequently unclear what else, if anything, they have here.
I do agree "background" is vague, but I generally assume the minimal threshold to say that is something like a previous career, undergraduate degree, or graduate degree in the background. I personally wouldn't say I had a background in something I was only an amateur enthusiast of.
---
That being said, what on earth gives you the idea a substantial background in sociology WOULDN'T be relevant here?
Because I understand what sociology is and isn't, since I too have "a background" in sociology and urban development. As well as statistics, history and even futurology.
Appeal to authority is a bias for a reason. In this case it wasn't even that well executed.
A "background in software development" (an HR or a marketing girl in the website design sweatshop) does not give enough relevance and authority to speak about the technical details of Linux kernel. Vice versa.
With his infamous "As I remember London" article [1] he directly correlated skin colour to nationality leading to the claim that native Brits had dropped from 60% to a third.
If you check now using current sources the British-born population is still around 59% the only thing that has changed is that fewer people identify themselves on the census as 'White British' which has a host of reasons around it. So he took a misleading statistic, interpreted it incorrectly and then built some kind of nostalgic story around it to support what was essentially an excuse to support far-right thugs protesting.
>So he took a misleading statistic, interpreted it incorrectly and then built some kind of nostalgic story around it to support what was essentially an excuse to support far-right thugs protesting.
His exact words from the blog:
"In 2000, more than sixty percent of the city were native Brits. By 2024, that had dropped to about a third."
White Brits are the only native/indigenous Brits to London. It's not misleading, it's just a fact. He did not interpret this fact incorrectly.
That's disingenuous because it's a neologism based on the population that were in America at the time of its colonisation. Native literally has the etymology of place of birth in the word.
You're mixing nationality, a piece of paper, with being "native", an ethnic group. Does not work in countries across Europe, Japan, Korea, etc. The settling of the USA and its cultural dominance has broken peoples brains. Europe is inhabited by its indigenous people!
The English are an ethnic group (ask 23andMe) and they are the founding stock of Great Britain. Source - the far right extremist wikipedia: https://en.wikipedia.org/wiki/English_people
I don’t actually believe this statistic, as an aside, but if we stipulate that it is true, it still doesn’t remotely mean 97.7% “ethnic English”. That’s a radically wrong implication.
It’s a little closer to DHH’s point about the percentage of British people (the Welsh, Scots and Northern Irish are not simply “ethnic English”).
But it’s still very wrong. At that point it was a ninteeen-hundred-year old working international port city, it has had a large population of Jews of European descent for centuries (the first large community being so notable that they were discriminated against in Magna Carta, after all), various displaced populations after WWI and WWII, a very large Irish population, thousands of Italians, many Dutch.
In the 1800s French Huguenot refugees and their immediate descendants made up perhaps ten per cent of the population of London (figures vary, considerably, could be 3% or nearer 15% depending on the boundary definition you choose and that at the time of their arrival the UK had no formal immigration policy so no exact numbers — but over 200,000 settled in mainland Britain). A truly astonishing number arrived and the city simply made room. But they didn’t leave; they assimilated. Same thing happened with Dutch exile craftsmen.
So he is wrong (and IMO ought to have known some of this) and you are not at all on the right track either.
People just have no grasp of the nearly unique reality of London as a historical melting pot. By 1960 its diversity was eclipsed by American cities like New York, Chicago and Philadelphia, but this idea that London was somehow traditionally “ethnically British” to the point of being a quaint monoculture is fully crackers. Just not true. The difference since 1960 is just more visible because we started adding more literal colour from the Commonwealth.
A place worth visiting for Brits who don't believe this and tourists who are interested:
The literal definition and etymology of 'native' relates to birth, check a dictionary and the definition will read the same. The truth is that 60% of London's population is made up of native Brits and that hasn't changed at all over the last 25 years.
You think Celtic Brits don't count as white? Have you seen us? We're some of the pastiest looking people you will ever meet. I could get sunburnt opening the fridge.
That is also bullshit (although a popular myth, deliberately cultivated by 19th-century English supremacists to justify subjugating their Celtic neighbors). Every region in Britain has a majority of Insular Celtic DNA on average (though the fraction varies). Even in the most heavily Anglo-Saxon regions like East Anglia, Germanic DNA is in the minority.
And of course the Celtic Britons replaced the Neolithic farmers, who replaced the Mesolithic foragers, and so on. "Indigenous" becomes meaningless with so many population replacements.
Thank you for mentioning the effect by name. I read first about it 20 years ago, memorized it under a totally different name, and was never able to find it again.
> The Gell-Mann amnesia effect is a cognitive bias where a person spots major errors in a media report on a topic they know well, yet completely forgets that unreliability and trusts the rest of the same publication on unfamiliar topics.
Yes. I am broadening the definition of "media" and "publication", maybe, though it is posts on his own publishing platform and performances at his own breakaway conference.
But if someone misrepresents deliberately on a topic of polarising importance, why would you spend any time listening to them on any other topic?
I would (still) listen to him if I did not know he wrote stupid wrong things on other topics in bad faith.
Do you automatically dismiss everything someone says based on their opinions on other topics not relevant to their main job? Is everything always absolutes?
DHH or anyone else spouting opinions on social media is hardly qualified to be setting policies for countries. Social media is also hardly the place to have any sort of debate on those topics.
Aha, the common HN tactic of expanding someone’s argument until it appears ridiculous and then attacking that. But I will bite.
It tends to only be when someone is egregiously wrong or (as in the case IMO) acting in deliberate bad faith on a topic of massive public concern, like this. Because these people are a waste of psychological resources.
I totally write off proselytising antivaxxers who I think should have reason to know better, I tend to write off climate change deniers who should know better, and yes, I tend to write off educated people who mislead with statistics to create intellectual cover for xenophobic, bigoted thugs and their antisocial campaigns.
This is not an exhaustive list, but it’s a good one. Other people who get written off tend to have a nasty streak on a topic of some divisiveness. For example I am not religious but not anti-religious; I am anti being deliberately or loudly cruel, posturing or nasty about it either way. People who are like that can be like that, but they will not get to tell me about their professional interests as if their awful behaviour is just a quirk.
YMMV, but I am now middle aged and I don’t have the time for this anymore. Freedom of speech isn’t without consequences for loss of respect. If you think my position awful or lily-livered you are fully within your rights to let it colour your judgement of my opinions on technology.
Apparently somewhere between a third to a half of American scientists believe in some kind of god. Not to denegrate those who believe, but that's some kind of compartmentalization going on. Politics isn't any different. I dunno, I guess I don't have much issue listening to people with whom I disagree. Some people just get stuck on things. I do to, and maybe I'm the one that's wrong.
> That surprises me, and I suspect it is not a good omen for Rails.
I had a fairly opposite reaction: Knowing what you can't be and how much less important your past work is going to be is good. If there are good reasons for more people doing less in Rails, then the creator is probably the person you want to be most level headed about that.
I literally just patched an AI-generated API endpoint on a new service at work that shipped without any auth whatsoever, because AI was re-implementing the auth token check method individually in each child controller instead of implementing once in a before_action hook. That's Rails 101 stuff and the app is small. It was so obvious that I saw it right away just reading the code, I didn't even set an agent loose to do an initial inspection. It was a real "yeah I still got it" moment for me.
I'm on board with the idea that agents are going to write most of the code, but not checking it is just insane to me, based on some of the things I've seen committed in commercial codebases recently.
One of the strengths of Rails was that it helped developers who maybe didn't always know the "best way" to build things to build in a standards compliant way – not that you couldn't do things like return a 200 for Not Found if you really wanted to etc.
I'm always pleased when working with a dev who has Rails experience even if they've since moved stack, as often they have picked up and learned the value of these good behaviours. Perhaps it's as much a community thing as a framework thing, or combination of both.
Agentic development definitely still has room to improve, but I suspect that there is some future framework (maybe even language) designed to help minimise these types of mistakes.
Interesting that all the patterns Rails adopted for humans to write better code can't easily be funneled into the LLM doing the job now. Seems like it should be.
I think the issue with Rails, as a long time developer in that ecosystem, is that much of the default assumptions are a) not common (not so far off-piste that a human can't get used to them, but outside the center of the bell curve), and b) not explicit. So it's very common to do things like install a plugin that has across-the-repo effects but that is only documented in e.g. the config directory and the bundle. Unlike, say, NPM, when you install a new package where you explicitly have to e.g. import it everywhere you want to use it.
As time has gone on Rails has gotten better about this, but it still means that "your rails" and "my rails" are different enough that AI could easily get confused.
As you say, the other aspect is the intense dynamism which works against AI, which again prefers to have everything explicit and reasonable about from the text.
So I ended up going back to Javascript after a brief 12 year digression into Ruby. I still love Ruby but it's just not the right language for the current environment and team structures - typescript is "worse is better" in the right ways that it's a much better choice for both frontend/backend being the same and better type systems.
Types are great. No nulls are great. Exhaustive switches are great. Derive macros and Serde are great. Detailed errors are great.
However Rust has one serious problem, BUILDS. It’s slow and it takes up a ton of space and memory. You will be working around this on any sizeable codebase. Using cargo check to ration your compiles, a dedicated build box, etc. Now imagine you're using it for a web application with compiled maud templates... you're potentially waiting minutes to see your changes.
If we accept that over half of the people building with rust are using agents (I'm sure this is much higher in reality), then you should also accept that there is a ton of time and resources spent on building / compiling. Optimizing this should probably be a top priority. They need someone who will be fanatical about shaving time and memory use from the build process.
Sounds like OCaml has the answer then. It has the equivalent of all the great things you described in Rust, plus compile speeds similar to or better than Go. Agents should be able to blaze through changes with it.
Yeah. I'm going through an exercise right now with a large codebase where I'm split some of it up into crates so that iterating on specific parts incurs less build overhead. Literally changing my architecture... which annoyed me a bit, but it does make sense.
> Agentic development definitely still has room to improve, but I suspect that there is some future framework (maybe even language) designed to help minimise these types of mistakes.
Maybe there will be one day, but what are the odds that it will meaningfully move the needle for existing platforms such that manual code review is unnecessary? "Blindly trust the output" is not a feasible security posture today.
I still review PRs, but rarely suggest changes. The most meaningful reviews come from our review bots. I mostly review broad architectural decisions as a way to keep abreast of changes in the codebase. There's a cohort of engineers I work with who I would be perfectly okay with letting the clankers review, approve, and merge their PRs. But there's a larger cohort of engineers who need what I would call a directional code review.
Agile is dead. It presumed writing the code was the slowest part of the cycle. Now it makes more sense to only start writing code when the requirements are known as the code is quick and low cost to change if/when the requirements change later.
But he himself promotes such a future too. This is a conflict of interest.
You can say "ride the wave", but this is more - this is helping the wave become huge. And cashing in too. Sorry, but I no longer buy any of what these guys do. To me they seem like BigBro ad-clowns.
> Agents do currently work well with Rails
According to whom? And not everyone wants to become dependent on corporate-agents sucking them dry financially. We all know the increase-in-cost will come with regards to tokens, after the get-hook addiction phase is over.
Nothing that he said is just him espousing the truth. It is the world as he and all the other AI-pilled people wish it to be. It is a world they want to convince everyone is inevitable. There are still a lot of people outside of tech bubbles that don't like the direction AI is going and don't want to embrace what these companies are doing or how the government is handling it.
And honestly, do people really think everyone is just going to become an "energized creator"?
AI is not as big of a productivity boost as compilers were but it is not zero. Back then, SQL was created and marketed as something “managers could use”, another parallel to the current cycle. This whole hype cycle is something we’ve seen before and IMO, we should know better how to handle it this time around. It’s a bit disappointing to hear the same arguments every time.
This has nothing to do with programming or productivity. I am addressing the fact that people like DHH and his fans are in the habit of framing the future as written and that you can either embrace it or get crushed. Embracing it is giving OpenAI, Anthropic, et. al your approval to keep on going. I'm not denying that AI can do the job a lot faster.
Keeping the comparison to compilers, people in the 60s refusing the transition weren’t crushed, their skills were devalued. The open models work well enough at coding to keep up without buying OpenAI or Anthropic products.
Given that we know he's literally a fascist (it's in all his political posts) that jives. "Embrace what we want or get mercilessly crushed" is a very fascist position.
Nah I'm good thanks. Call me old school, but I'd rather understand how things work, and try and keep a maintainable codebase. Rather than just get some random rust I've no idea what it is, or why it does it, pushed into prod.
I think this is going to be todays 'cheap outsource' culture. Good enough to keep shareholders happy, an apparent boost in productivity, before tanking the reliability of the company.
Remember when we used to scorn stackoverflow developers, or PRs that were more than a few lines long. An every morning git pull is now bigger than my terminal buffer, its complete chaos. Numbers game to keep the new metrics world gamification going. Its dire.
1. AI is exposing the harsh truth that 99% of the code out there has always been throwaway code. Yet good developers, void of any engineering skill, treated it like the 1% that wasn't and called their hygiene and practices engineering.
2. Way too many developers that have made their careers or internet fame out of their technical proficiency understand that the shift is unavoidable. See Antirez implementation of Arrays in Redis or his Dwarfstar project. Virtually every line of these projects has been AI authored. And yet they are terrific projects with extremely strong architecture. And not a single line was edited in vi (his favorite editor).
> Virtually every line of these projects has been AI authored
There's a difference between everything being vibe-coded where you don't look at the code and don't understand anything and using your agent like a very fast pair programmer.
> In the first month I just wrote the specification document...Thanks to AI, the specification evolved a lot, via back and forth of feedback, intellectual challenges about what was the best design, what was the right compromise, what was too engineered and what not.
> You know what was the biggest realization of all that? For high quality system programming tasks you have to still be fully involved
Judgement will be required even moreso now. There will be critical parts of the system where it's important that engineers know them well. But there's a lot of work around the core that isn't that critical to know in detail. For these types of things, I've found it's just important to understand the gist to make sure things aren't going off the rails. But for the core pieces, it's important to understand the details of what's happening.
I agree that we have to keep understanding our code and we have to understand the systems we build.
I think the deeper issue is that AI can generate so much code that reading it simply doesn't scale.
Even if AI were to produce perfectly written code that didn't have any bugs that human review needed to catch, it would still be too much code to read in order to understand everything. The machine can write much faster than I can read.
Extending this argument further, it also becomes impossible to onboard young people to codebases if we maintain an expectation that reading all the lower level source code is how they come up to speed. I don't think this was ever an actual expectation in the industry, although we may have short-handed it that way in conversation.
So IMHO we need to come up with higher level visualizations and other ways to compress the knowledge that we used to get out of reading source code. We basically need to create new higher level projections off of our semantic understanding of the project.
A lot can (and should) be said about this keynote, but I have a hangup on portraiture as an example of "creative destruction" (how I loathe that expression!). Classical portraiture is still alive and well and artists are still able to make a living from it - possibly even more of them now than during the 1800s.
The advent of the camera did push visual art in new directions, but not quite as presented here. Oil painting and portraiture wasn't as static as it's made out to be, and there's a difference between a commissioned portrait and artists painting people because they enjoy painting people. Giuseppe Arcimboldo, El Greco and Hieronymus Bosch are some of the most well-known experimenters in western art and they were active in the 1500s. Some of the most famous art movements (impressionism, expressionism, arts & crafts) appeared well before photography was democratized.
The biggest difference, though - as much as I love art - is that portraits are a distraction. Then as now, it's mostly a vanity project for the affluent and of little consequence to the lives of the vast majority. Software, on the other hand, is not only a multi-billion dollar industry employing millions of people, it's also a direct matter of life and death, since it's an essential component in things like nuclear power plants, airplanes and medical equipment.
Hieronymus Bosch is such an interesting case, and I like your point in comparing oil painting to photography. Unfortunately I also share your concern about software's criticality. Therac-25 comes to mind as a case study taught to us in computer science university courses, that have haunted me in recent times: https://en.wikipedia.org/wiki/Therac-25?wprov=sfla1
> "creative destruction" (how I loathe that expression!)
It is sadly a fundamental feature of capitalism. We will see this again and again as long as our economic system is set up the way it current is. Just imagine, there will be a time in the perhaps not-too-distant future when the old folks of that time will lament how easy the kids of today have it!
So the keynote mentions that in a perspective you are going to be a "maker of things" rather than a coder. But, nobody is asking, in a perspective, why would anybody use the stuff you "make" rather than using AI directly? All of the notion of apps disappears, in a perspective.
Do you change the oil on your car yourself? No? Why not? Buying the oil, filters and all the tools you need to do it forever would cost less than one oil change at the dealership. Plenty of videos on YouTube you can learn from.
Sure, but changing the oil means buying a bunch of stuff and getting dirty. Replacing your vibe coded garbage app with my vibe coded garbage app is much much easier than changing my oil.
That said, I think this only applies to apps that can be replaced by vibe coded garbage. Anything with a network effect, or anything that requires deep domain knowledge is much safer
> Replacing your vibe coded garbage app with my vibe coded garbage app is much much easier than changing my oil
I am not sure what is it about the HN crowd that makes them so out of touch with what the average Joe is capable of. It's like that Dropbox post top comment all over again.
Don't you think that the only reason you think it's simple is because you're... A professional Software Engineer??
I know for a fact that it has gotten simpler. Which means that the number of people who can do a version of the thing i’ve been doing most of my life got way way way bigger in the last couple years.
Sure, not everybody can do it - i take your point. But the dynamic between can and can’t has changed. Projects I wouldn’t take on because they were too much work for not enough reward have become something I can pop off in a couple hours. People who know nothing about coding but a lot about their specific need can build complex apps in days.
> Which means that the number of people who can do a version of the thing i’ve been doing most of my life got way way way bigger in the last couple years.
Which invariably means that businesses who incorrectly believe that their vibe coded solution will withstand the trials and tribulations of production will be needing even more professional help when it all goes wrong.
Yes people shouldn't discuss until they have tried vibe coding as SaaS. I do this now and it is 10x faster at coding and 1x on any other tasks. So overall maybe 3 months work is done in a month. This means the typical SaaS may be cheaper now but it is not like anyone can vibe code it cheaper than the subscription fee.
> Buying the oil, filters and all the tools you need to do it forever would cost less than one oil change at the dealership.
Are you changing your own oil? The cost of oil alone usually ends up being around 3 DIY changes vs one at the dealer ship. The mobile 1 synthetic I like is currently $92 for a 3-pack of 5qt containers. Which is enough for 3 oil changes with some left over to almost do a 4th. Dealership currently wants $84+tax for change w/free tire rotation. It's cheaper but certainly not "do it forever", and living in an apartment often means risking getting a warning from the landlord for doing auto maintenance in the parking lot since its against the stated rules in the lease.
Drain plug is in a stupid location that I don't often feel like dealing with. I do change the oil myself in all my other machines with much more pleasant for oil change designs, though. So I guess what you are saying is that AI isn't ergonomic enough yet, but will that hold true much longer?
I have been thinking quite a bit about this lately. I am in the process of building my own music software applications. This is actually allowing me to leave MacOS since Ableton was previously the anchor for me. I can now build tools that are more geared towards what I always wanted in Rust much like DHH was talking about in that talk.
I could sell some of my tools I built but I am on the fence about it. As a few others have said, there is still work and taste that goes into those tools that some might not want to put the effort into. However, there will be an increasing number of people who just build bespoke tools for themselves. That's probably ok. There's a market for both.
The most important thing now (and maybe always) is distribution. If anyone can build something then it ultimately comes down to your ability to market, build trust and provide support at 2:30 AM. That's hard to do when you're a one person shop. Software is an ongoing operational commitment if you make it your business.
Because it does take some skills, judgement, taste, knowledge, time and effort to build things. He said it is not very different from directing a team of contractors, and also said that a 10x dev in now 1000x with a tool like this. He also gave the parallel of photography to address this point, that people were forced to be more creative and adapt.
But I assume you didn't use to build anything substantial.
Business people I work with tend to have all kind of ideas in different places, I cannot count all the Trello boards they have abandoned. Users are usually lazy and don't want to fill in data properly, well yeah chatbox with llm is cool interface, but you still have to enforce structure for data so it is useful.
With AI problem is still there "garbage in, garbage out", models are not going to imagine correct data to fill in blanks, in the end if you need someones birthday, someone has to type it in.
Web app I am working on is organised only because we get engineers and quality assurance and we get paid for maintaining application. Data is structured and we can deliver insights because the application enforces structure.
There are far more requirements than AI can deliver on its own.
At least at the current level of AI, it still takes days/weeks/months to make a good app if you have standards. That's not effort everyone will want to do.
If/when agents become reliable enough, there will be no need for web apps, nor phone apps, etc. In fact, DHH got that part right: "your app should have a CLI if you want me to use it." If "English is the best programming language", then English is the best UX. I'm not navigating a silly page and filling out forms and dismissing popups if my agent can book a flight for me with a sentence and a bit of textual back and forth.
I wonder if the web will go away and the Internet will be just bot to bot traffic.
Navigating the web seems archaic in that context. Not only is that the end of Rails, but of frontend, JavaScript, etc. in the long term.
So far a Rust rewrite's success rate directly, probably super linearly, correlates with the presence of a good, well designed test suite. This is why Bun's rewrite and others are successful. I anticipate that this will be true for the 37Signals rewrites.
I want to learn a couple of things once they are done:
1. Once the test suite assistance is over, how do features get built in their backends and continue to maintain the same quality. LLM test writing is a bit trickier as it often favors what Randy Coulman called tautological tests back in the day. It takes a lot of work to get them not to do this, so I am guessing folks at 37Signals will read at least this part. I can't imagine zero Rust read at all.
2. Over the years they had a few innovations in their domain design, ones that Rails made easier to do, for instance their delegated type pattern. Will Rust replicate this, and how would one know without reading it, or perhaps it does not matter? Furthermore, how do newer patterns emerge? and how does our arsenal of better abstractions keep growing?
I don't have the answers but I am glad there is a chance we can learn said answers.
Classic DHH.Allways talks about what is good for his situation, not what is good for the community in general. I can understand his excitement of slashing his server and labor expenses. He will be richer and have more time, less people will work for him and wow, rails came a long way, didn't it ? The world is your oyster now! Oh and btw, we are changing the name of the conference, it's called Rust conf now.
It is quite telling how virtual no conference talks mention concepts like solidarity or technology that is useful for society as a whole (not tech that just enriches the few).
Regarding the native vs web app problem, this was a completely made up human-made problem. I think I would have preferred a deterministic solution like a game engine but having agents code up native apps based on your web spec is not a horrible idea.
The bigger question to me, is the original implementation, are agents going to do that, or will humans hand code it. My two cents, I’ve been using agents heavily now, I review thousands of lines of code a day, I don’t think this works unless I have a clear idea of the spec I want to build in my mind already.
I get a better understanding of what the architecture should be and how I should build it by actually doing the thing, which means typing code at human speeds. I don’t reason my way into a good architecture, I don’t think I ever did that much in the past and I definitely don’t get it by asking the agent to check its plan and explain it to me the hundredth time. And then once I know the architecture, I don’t think the agent even speeds me up that much. It changes 7 days of coding to 1, but those 7 days weren’t worth much and I generate even more ideas and mental models in those 7 days that pay dividends for years to come.
Would you have told a tailor that using the industrial sewing machine improves his skill at making delicate, beautiful clothes? Of course not, actually even in 2026, large swathes of high quality delicate clothes are hand sewn. There will be an equivalent for coding, even DHH admits it, when he hedges with “most”.
I think, for me I’ll navigate my career to focus more on extremely high quality, delicate custom code and leave this agentic coding to the sweatshop workers. I wanted to do this anyway before agents, now I need to accelerate it. yes FAANG is sweatshop work.
I was hoping there will be news on Rails 9.0. Feels like this is the end of Ruby Rails.
Hey is not even a Ruby Rails App anymore. And indirectly admitting Ruby is slow considering they manage to cut 99.9% of CPU with their rewrite in Rust.
I hadn't connected the dots until you mentioned it, but yeah this keynote indirectly says it's the end of Ruby and Rails. Ruby was designed to be coded by humans, and DHH says it's the end of coding by hand.
Even the concept of "framework" as we knew it becomes different in this age. If you only want reusability, why not use plain good old libraries and let the AI assemble them as needed for your project?
> Hey is not even a Ruby Rails App anymore. And indirectly admitting Ruby is slow considering they manage to cut 99.9% of CPU with their rewrite in Rust.
Let's see how that goes.
Personally, I wouldn't want all my emails stored on a service where the dude doesn't even look at the code anymore.
Those performance numbers mean nothing considering they went from rendering every single view of a highly interactive app solely on their servers to moving all rendering to the clients. It’s possible the backend logic of the app could have run on 10 hosts running Ruby as well.
I agree it seems like the end of Ruby on Rails. I personally hope a fork will come along as a spiritual successor, because I still love building web apps this way.
And Shopify has been sponsoring Ruby JIT developments, exactly for the same reasons Facebook created their own PHP fork with a JIT, and how Twitter moved into the JVM after so many fail whales.
Unfortunately yes because Rails has always been a framework of extraction. Now they are not even coding anymore there is nothing to extract. No one is a programmer but everyone is a maker.
I actually believe new frameworks and languages will pop up that are explicitly made to supercharge agentic engineering workflows. We're still super duper extra early! November last year was the START of all this. We're still in September! Not even one year has passed!
It doesn't make much sense to use a human language like Ruby and leave so much on the table when you can use a language like Rust or C to squeeze a LOT more out of correctness/types/performance.
This is a question I have been circling around. The agents are good at Ruby/Rails. It’s probably a confluence of things, but whether the L in LLM benefits from Ruby’s emphasis on human readability is something I’d love to see a study on. Or more generally, how key is (human) language itself in the performance ceiling of LLMs. A machine specific coding language/framework/process could obviously be more efficient, but if language itself is the hard limit, those would be dead in the water with LLMs. Or put another way: if language wasn’t the ceiling, why didn’t we see machine first coding implementations before LLMs?
We're going to have a bunch of hardly used libraries/frameworks that might converge on a single one. Similar to the surge in JS frameworks/libraries from a few years back and where we are now.
Is squeezing a lot more performance beneficial to a business? or is it a small amount of VPS cost per month difference?
It helps to be clear, at least with yourself, about if you're doing something because you can/want to or if there are actual business reasons.
I just skipped to a random part, 41:47, where the speaker is talking about the chasm of the old and the new, the before and the after, and our acceptance of the new status quo.
These match my own thoughts exactly. As much as I don't love them, I agree with the speaker.
For Factorio players: we used to live in Nauvis. It was beautiful but sparse. Then we outgrew it and unwittingly found ourselves on Fulgora. It is abundant but trashy.
And just like in a game, the rules changed and we adapt. That's what makes the game fun. Granted, that it's all our livelihoods on the line (very high stakes), but all things being equal, we have new challenges in front of us that we can either bemoan or accept.
Where did he say that? I listened to it yesterday and AFAIK he only said "I don't know where the line between native vs web will end up in the future" (now that we can generate native apps easier)
Keep in mind DHH is working on an email app, it makes sense he sees native as the default option for that over pushing everything through the web when the barrier to entry lowered dramatically.
Hotwire is not explicitly a cross-platform concept. It's primarily a way to build web apps in the browser which is not going away any time soon. The only difference is in 2026 you're just more likely to vibecode a native mobile app than build a web/React Native version. Likewise you may build a native Desktop version instead of an Electron wrapper.
The issue with native apps is that you have to convince people to install them. If I had one app for each tab I open in my browser my phone would be a mess and out of free space. Then you slow down the delivery of features and bug fixes because you have to go through the app stores. You can't deploy something every day or more often.
So even if we can generate an app easily now, is that a wise choice for every business?
Like everything, depends on the usecase. An email desktop app seems to be a valid one. A standard earlystage B2B SaaS Rails web app? Not so much.
Same with server side rust. A high volume transactional backend email service makes sense there. The base layer of a CRM managing a lot of business logic?
Life will go on much the same. We just have more options to say yes to 'native' with LLMs, we can experiment more with ideal scenarios rather than a religious adherence to cross platform or single-framework merely for team DX simplicity.
If the apps were actually good, you'd install them. A lot of the apps on F-Droid are pretty lightweight, and a phone's security model makes it much safer than installing apps on Windows.
The web makes apps worse by forcing a fairly strict sandbox, even as Google keeps trying to loosen it. Think of a messaging app that hides in your notification tray, that's a trivial UX thing that's impossible on the web.
When I started in computing there was no Web, there was enough space to install the applications that actually mattered, and even better, developers actually had to take hardware constraints into account.
Rendering HTML on the server was kind of how IBM mainframes talked to their 3270 terminals, only that HTML is a lot nicer than 3270 streams and HTTP is human readable.
Server side rendering makes sense for mostly static content, but for web applications, a front end framework and a back-end API are a much better match.
I really enjoyed DHH's keynote speech using his own family lineage story via the art history intertwined with technology evolution portrait paintings -> photography pictures as analogy of what's happening right now in the technology landscapes.
I am glad that he mentioned that he'll miss the flow state that he would enjoy while deeply focused in computer programming. That was also one of my worries. Gladly, I still have other ways to get there, and I hope that you'll find yours too.
Out of curiosity, what other ways do you have to reach flow state? I honestly struggle to name even one. And definitely not one that would pay the bills.
Well, an email server is next to the best use case for a low-level language. I'm still choosing Rails (or Laravel or Django) for CRUD apps any day of the week. Rails in particular is great because it's very conventional and AI writes it in an standardised way.
I don't understand what is supposed to come next. So its not Ruby on Rails anymore. What is it? Are agents going to magically convert all the RoR ecosystem to Rust or Whatever?
But yes, I think RoR's selling feature was developer ergonomics, which suddenly seems less of a benefit if developers aren't the ones writing the code. Readability and conciseness still has benefit, but the trade-off of worse performance (and less static checking) is suddenly more questionable.
> Readability and conciseness still has benefit, but the trade-off of worse performance (and less static checking) is suddenly more questionable.
There is an issue with this idea - we write software, we read it, we make sure it does what it should do, that it’s implement in a sensible and efficient way, and, most importantly, we make sure the tests do what they should do.
If you no longer write the code, it’s harder to understand it. If you don’t write the tests, you need to trust the LLM did both jobs correctly (multiple adversarial LLMs can help), but, ultimately, it’s code you can no longer claim you understand. And you can’t really be sure it does what you want until you understand it.
Shit, I'm also hosting with them. Somehow I missed that post. I've been watching Kurt's AI-pillness advancing over time, but I didn't expect it to metastasize so quickly.
What a strange time. Kurt himself writes that they've had the best quarter ever, and yet he feels confident enough to steer the ship in a completely new direction.
Write a solid piece of software by hand. Spend 20 plus years working all the bugs out and perfecting it. Build a giant community around it. Then use llms to rewrite it in another language, as the last step. Once that becomes unmaintainable, and there really is no community you can start over again by writing another working piece of software in a new language, perfecting it, building out a large community and then at the final step translate again with an llm, repeat.
I think what's next is the harness. Ruby on Rails brought conventions to web development. It could morph into brining conventions to building any kind of software using AI. AI on Rails. Many such efforts already exist and perhaps in short order such a project would be obsolete, but it's one potential path forward.
I think the perf jump from Ruby to Rust is much greater than .NET Core to Rust. Not much will change in .NET land for the foreseeable future. I see this migration happening from Ruby and others so it is not surprising from that aspect.
Also - what he is talking about was true for well over a year. It is just that a lot of day-to-day gaps between prompt and output were closed in the meantime. In a sense, Ruby on Rails/37 Signals is a late adopter.
> I think the perf jump from Ruby to Rust is much greater than .NET Core to Rust
This has not been true for me. If you are constrained by memory, you will still come out better using Rust, even if you use C# with AOT. We use .NET 10 at work and see 1/5th of the memory being used by a ported Rust service. Your mileage may vary.
I stopped writing code manually more than a year ago. I've experienced the 100x, even perhaps the 1000x he mentions, and I still think he's fucking psychotic.
Not reading generated webapp code is one thing. Blindly commiting OS code is a whole different dimension, as is writing production code in a language you don't even know how to read.
Worry not, for I am using Claude right now. I would never question the doctrine of AI usefulness and productivity. Please scrap that J'accuse draft, sir, for I would never commit blasphemy against AI, I promise.
It's still early days, I'm hesitant to claim the death of software engineering. It's good now because LLMs were all trained on actual human output. One possible future is a slop death spiral.
I sure as hell am never again releasing human-authored open-source in a way that can be scraped. My Github is now relegated to the lowest quality slop from the cheapest models.
Serious question: Is there a framework that is better suited for agents? There must be something out there that makes agents fly through development with minimal token expense.
Yep I find Elixir very enjoyable with agents at the moment, I've been developing a lot in it. It helps that I already liked the language and the platform, and now I just have the ability to build out my ideas faster.
I think they're good at it for a variety of reasons, but it likely helps having a smaller and newer community than many languages, meaning less out of date guff clogging up the LLM sources.
I maintain Elixir and Rails applications. In my opinion, almost all LLMs handle Elixir better. I suspect, though do not know, being an immutable functional language and composing applications in a functional style makes it easier for the models to "reason" about the code and state. I learned very quickly that this is my preferred language for that very reason.
Additionally, the features available in the BEAM and OTP mean the agents have built-in tools for many things that would require reaching for 3rd-party libraries or even additional services. This means less variability in design as the models will happily use all the language features human developers would have to learn about and understand over time to use effectively. Not that this isn't important, but I believe models do a better job steering you to obvious and optimal solutions with Elixir.
The new type system is only going to make it better.
An agent that was looking at the log output from a program in a typical language, and also looking at the code for that program, would have a lot of unanswered questions related to "how do I iterate on this?" Maybe there's a kubernetes cluster involved somewhere. Maybe it there's a cloud storage bucket. Maybe there's dashboards and metrics and telemetry hosted by servers that the agent doesn't know the name of.
Elixir runs on the BEAM which lets you do hot code reloading process at a time. And these are much more lightweight than your typical OS process. So you can do things on the BEAM which would be just crazy to do on other achitectures. Like if you have a server which needs to support a million users, it would be insanity to run a million separate OS processes for the job. But BEAM processes are so lightweight that you can do that, which gives each one isolation from misbehavior of the others.
And since these support hot code reloading, without restarting the BEAM's OS process, you can patch the code for your user's sliver of the server and do a sort of "what-if" experiment. Other architectures would require you to implement separate test environments or complicated feature flag systems, but the beam lets you just reach out and change it for just the process you care about--you're not baking the experiment into an OSI image or anything so dangerous as that... blast radius is kept small.
And the steps for carrying out this experiment, it's all doable at the elixir shell. You connect to the BEAM and make changes. That's a tremendous reduction in context that an agent needs to load--context which would otherwise instruct the agent about how to reason about the deploy process and how to wall off your experiment from others so as to not cause problems with it is just answered implicitly by how the BEAM works.
Elixir compiles to BEAM bytecode, so you get all of this because the BEAM is cool. As for Elixir as a language... I don't think there's anything agent-specific about it. Gleam might be a better choice since it's statically typed and also compiles to BEAM bytecode except that it's not as popular as Elixir, which is probably more important.
You can also do all of this in the language that started it all: Erlang (Java : JVM :: Erlang : BEAM) but it's not exactly fashionable at present (similar to java vs like... kotlin or clojure or something).
Disclaimer: I'm more of a fan than a seasoned pro. I'd love to actually be working with this stuff daily, but my team wouldn't have it because they prefer familiar things and pain.
Also, he always states super strong opinions on anything, punctuated by the f-word, as if everyone else was shit, and then he changes his mind. (For instance, the way he despised VCs and now he seems to acknowledge them, his stance on AI, …, his change from being seemingly more left-leaning to being a libertarian).
He several negative traits. He's one of these opinionated controversial guys. Because he has money people feel the need to react to what he says, whether they think he's right or wrong. Annoyingly, people like that get more exposure than level headed people with measured takes.
Mixed with his terrible views on society (xenophobia, techno-fascism), it makes a rather sour cocktail.
"Strong opinions, loosely held" is an ideal worth striving for. The alternative is to hold on to things you should know better are false, aka dogmatism.
One thing I really agree with dhh on (and something I’m genuinely looking forward to!) is the resurgence of native apps.
Cross-platform became so dominant because the same small dev teams had to support users across multiple platforms. Practically they either had to build ONLY for the web or use some kind of web-tech-based framework like Electron.
Agentic coding changes that constraint at a very low level. Teams are no longer limited by whether they have someone who knows how to build for iOS, Android, macOS, etc., or whether they have the bandwidth to maintain multiple native codebases. The cost of supporting multiple platforms starts to fall dramatically.
And I think this is closely related to the thesis behind Omarchy also. Setting up and customizing Linux is historically a lot of friction. If you wanted a system that was personal i.e. NOT (Mac OR PC) you often had to be comfortable troubleshooting obscure issues and spending hours getting everything just right. Don't get me wrong: a lot of people enjoy that friction. But it sucks at so many levels.
Agents lower that barrier. You can want a highly customized system without necessarily becoming an expert in every layer of the stack. If there’s some tweak or workflow that isn’t exposed through a settings menu, you can increasingly just describe what you want and have the system help you build it.
So I think the interesting shift is that software becomes less constrained by what the average team or individual programmer has the bandwidth to build and maintain. That could make both native apps and highly personalized computing environments much more viable again instead of the web-slop that we've become accustomed to over the past 20 years. I really see parallels between how the Mac empowered creatives in the 80s and agentic coding today.
As much as i'd love for this to be true, it is not. Teams need to support their native apps, just because you prompted your way into some code doesn't mean your team is able to support it. The constraints are the same, if no one in your team knows how to build an maintain a native iOS app and their core knowledge is TypeScript and web, vibing it with agents won't get you past your first release (and not even that)
I don't really mean “give a team of TypeScript devs an agent and suddenly they can ship and maintain a production iOS app.” Expertise and ownership still matter, especially for something you have to support for years.
I'm talking more about agents as a force multiplier for teams who do have engineering experience. They reduce the amount of specialized knowledge, time, and coordination required to get something built and maintained. The constraint gets meaningfully weaker.
An experienced engineer who understands software architecture, debugging, product constraints, etc. but isn't an expert in every native platform can increasingly use an agent to bridge some of those gaps. That's very different from vibecoding your way into production.
fwiw I think we're still very early in the process of re-thinking about how teams are organized and what skill sets they need.
Totally disagree on the reevaluation of architecture best practices.
Agents are as prone as forgetting to update duplicated code as we are ( or almost ), because they work at the same level : text files spread everywhere that you have to open and read to know what's inside.
The fundamental reasons for those architecture best practices are still there, and it's actually quite convenient and also a good reason to remain optimistic in our ability to find a job as a professional software developer.
Once the rest of those talks are published on YouTube (usually in a couple of weeks), you’ll see that there were some other talks that were much better about conveying a nuance in a balance between the tension of “AI can do everything for me, let it write machine code” and a session titled something along the lines of “your code base is your prompt“.
One presentation, I think it was my favorite one, talked about how an LLM had a very difficult time understanding what happened in a class when the name of the class was poorly descriptive of that class behavior.
Good quality code still matters. The LLMs are based on language! Human language!
A lot has been said about DHH, but, I think the moment where you can see that the guy has _clearly_ lost it is this quote:
> It is impossibly fast. And you know what? The binary? Half a megabyte. Half a fucking megabyte. I— you could fit that on a floppy disk almost. With a little compression. This is the other payoff.
The guy now(?) thinks like a lossy abstraction that speaks in tropes.
He has become the LLM - I would say, if I didn't know that even a 9B one can tell me with absolutely certainty how much space there is on a fucking floppy.
There is no reason to get into all the politics stuff when this is where the guy is at. Basic fact retrieval.
(Pedantically, one could now argue that it was not clear which floppy format the guy was referring to. And yes, that would be true, but by thinking that thought, you have thought about the sentence more than its author has.)
If we are talking about the standard 3.5-inch High Density (HD) floppy disk, which is the universal cultural icon for "floppy disk" (and the save icon in your software), the capacity is exactly:
1,474,560 bytes
This is equal to 1,440 KB (using 1024-byte kilobytes).
It is marketed and universally referred to as 1.44 MB.
Now, let's run the exact math on DHH's keynote claim:
If you put his "impossibly fast" 0.5 MB binary onto a standard floppy disk, you would have nearly a full megabyte of free space left over. You could fit almost three of his binaries on a single disk. You do not need "a little compression." You need a basic understanding of less-than/greater-than operators.
The Preemptive "Reply-Guy" Shield
Now, just in case a Hacker News pedant is reading this and wants to argue that DHH was talking about a different format, here is the absolute certainty on the rest of the historical hardware:
3.5-inch Double Density (DD) (1980s): 720 KB. (A 0.5 MB file still fits easily without compression).
3.5-inch Extended Density (ED) (1987): 2.88 MB. (Fits five of his binaries).
5.25-inch High Density (1984): 1.2 MB. (Fits two of his binaries).
5.25-inch Double Density (1981): 360 KB.
8-inch Original IBM (1971): 80 KB.
To make DHH's statement mathematically correct, we have to pretend he was standing on a stage in 2024, hyping up modern web framework binaries, while indexing his storage metaphors to a 5.25-inch Double Density disk from 1981.
__
Thank you, Gemini. Much appreciated.
Also, I love the tone.
(You can tell that it is actually that LLM by it thinking that it is still 2024)
weird talk; is not reading any code _at all_ is the new standard now? I do happily ai-slop side projects here and there but seems like just going full steam ahead _at work_ is ok? sorry, I meant - is the future now? xd
No, not at all. Anyone who gives a damn at all about producing quality software is reading the code, and frankly most people who care about producing quality software aren't using LLMs to do it.
Haven't watched the video, but I've been spending a few hours the past week re-acquainting myself with Ruby before diving back into rails introductions.
Ruby is an interesting language to work with. I don't know it like Python, but it's a language that finds problems developers would run into and built solutions in to make it easier.
As for Rails, as a total novice: It is a lot of Convention over Configuration, and it's almost a second language on top of Ruby. Everything works together implicitly so you have to recognize what's happening when you look at the code. I don't know if any Ruby/Rails oriented IDEs exist that guide the developer with hints and shortcuts to how the pieces of a Rails app fit together, but it sure would be handy.
I can see how typed, high-performance languages with explicit references to all the integrations would be easier for AI to handle. But as long as humans sometimes have to look at the code, I would think Rails still has a place.
> how the pieces of a Rails app fit together, but it sure would be handy
Best thing you can do is get a fuzzy file search, but you'll learn the structure soon enough. Once you know it, you'll be very fast. I loved it when I used it professionally.
I sat front row for David's talk yesterday morning. Having talked with him the day before, I’ll admit… I wasn’t particularly surprised (nor unprepared).
A bit of a field report from #RailsWorld: the vibe here is far from doom and gloom. Quite the contrary. Wherever our industry is headed, most of us are still employed as menders… tending to systems that customers rely on and businesses are quite happy to keep paying for.
Most of us aren’t waking up to a blank canvas and designing the architecture of the future. We’re inheriting decisions made years ago, updating old patterns, working around constraints, and keeping this shit running reliably.
I think these newer tools give us an opportunity to wonder a little more about the systems we’ve inherited. To unpack why things work the way they do. To tinker with assumptions we haven’t had the time, confidence, or permission to revisit… and share what we learn so the next person, or agent, has an easier time.
That deployment model we picked eight years ago? Worth another look. Some of our web apps probably wish they were native apps. And there are plenty of architectural decisions we’ve been living with mostly because… well, we’ve been busy living with them.
There’s plenty of understandable anxiety about what these tools mean for our work. I’m increasingly curious about what they give us permission to revisit.
Once my keynote is published, I’ll share more about a little programming-language-adjacent framework I’ve been working on for approaching exactly this kind of curiosity.
In the meantime… keep showing up. Keep wondering.
Long live Ruby. Long live Rails.
p(bloom)
Its easy to be optimistic when you're sitting on millions (dhh). You literally don't need to care about how any of this affects your job security.
Right, or paying for AI models in general.
There is a world of difference of having tens of millions of dollars and paying hundreds of dollars a month for AI subscriptions vs someone making 100k / year doing the same.
For the DHH's of the world that's not even something to think about, but for most people who are working (even in tech), now we're talking 2-8%+ of your income depending how much you use AI. Just to have the "luxury" of being able to experiment with it. I use air quotes on luxury because I personally find a lot of AI results to be quite bad, but then again, I'm not spending hundreds of dollars a month on the premium models.
I make 55k working at a dental office and a restaurant and I'm paying 200$ a month on open source models because I'm having an absolute blast with them.
Plenty of jobs have expenses built into them. Maybe a subscription is more than it used to be, I had great jobs on a used 200$ thinkpad, but it's still cheaper than being a photographer. And 100k is more than most jobs.
Software dev, even if it's all agents now, is still pretty dang economical
Have you tried OpenRouter and the DeepSeek-like models? If you try to optimize a little you could probably slash those costs.
I send a ton of requests and even reaching $10 for the month would take more calls than I have capacity to review and use. And as I said, that's for a ton of requests.
I use Synthetic, Neuralwatt, and Opencode. My usage is fairly high, I'm at 1.5B tokens in the last 7 days. ;)
I'm paying $20 a month using opus 5.5 and having a blast.
20 a month and using sonnet mostly for random fun stuff.
At work, it seems a bit different - I'm certainly having to declaudify things it can be hard starting there's a lot to dig into - you can see it's overwritten but need to understand every detail to sort it out.
It's easier to enjoy it if you don't see how the sausage is made.
I am definitely not having a blast at work with LLMs. Everyone in the team has become simple rubber-stampers for whatever the AI puts out. There are mountains of documentation which gives a false sense of understanding of how things work, but the code repository itself could have been written in Brainfuck and it would make almost no difference. And we keep getting told by management that this is fine. We've become simple janitors in the slop hall.
1. Present a big AI PR to a fellow human for review 2. Understand a big AI PR, as a human reviewer
...is going to be a hero. This is one of the biggest unsolved problems of this new era.
I've been experimenting with this, but it's early days. My efforts are mostly along the lines of having AI generate artifacts that guide the reviewer through the PR in execution flow order, explaining the reason for changes along the way.
I don't think that's the way forward. We don't accept mega-PRs from people, why accept them from machines? The best practice of small, focused commits - whether it be from a person or a LLM - will win out in time once we're past this manic phase.
Yeah, I'm sorry to hear that. I'm very glad I am not dealing with anything even remotely corporate during this time, because I'd absolutely hate it
You think 2-8% is a problem? There's a large world outside the US. "Hundreds of dollars" per month is most of my salary, I can afford that only by going hungry half the time.
Probably going to if I want to keep my job, since I'll be making about the same amount of money moving boxes or cleaning toilets, only in much less comfortable conditions.
Yeah, I feel bad about those situations.
It's essentially trying to turn programming into a micro-transaction based pay to play / pay to win outcome, a concept that was gaming related for decades.
Meta just gave DHH $1m in credits fwiw
Can you expand on that? I couldn't find sources
I think it was for that linux script thing fwiw, omarchy
You can pay $5 for deepseek 4.1 flash and be on the pareto frontier. And you can pay between $20-200/month for a huge amount of usage of the absolute frontier models. I was concerned about costs a year ago but with the proliferation of great cheap models, I think it's no longer a concern at all. I'd be fine if I could only use 4.1 flash.
I think the fear, threat, and valuation, of the frontier labs is that they will cross some line and have models that are infinitely smarter and more capable than everyone else. If that should happen, your deepseek account would be meaningless.
But I also think that premise is bullshit and not gonna happen.
sure but if that's the case society and i will have much bigger problems anyways
The premium models are worth 10% of your income at least.
There is no future without these models. They truly are an entirely different league than the cheap stuff. You can move mountains with Astra or Opus 5.5.
10% of a software engineer salary is a lot of money. That's just ridiculous.
Like the other commentor said, its just going to be a net negative, because whatever productivity increase you supposedly do get, you will see exactly 0 of that translated into salary increase.
Have you tried to negotiate a 10% pay raise because “you can move mountains with Astra or Opus 5.5”? I’d wager you’re more likely to get a “the fuck do we need you for then” than that pay raise.
The whole AI issue is and always has been a social question. It's hardly a ridiculous concept to have a skilled worker control a $10k budget to buy tools. In many fields that is normal.
Previously, the tools in our field were free or trivial cost. Seems like that has changed, that's OK, there are models to solve that.
I teach a couple courses currently, there is an enormous gap between what my students can create with AI and what I can create with AI. I don't think it's that much smaller than the difference of what we can create without, the only difference is that the schedules are compressed.
They surely weren't, and piracy was quite common in many European countries, buying tapes and floppies from bazaars back in the days.
Let’s see the moved mountains I’m sitting here waiting to see any improvement in software
On the contrary, I say software got even worse
Even if you believe this surely it should be paid for by the employer not the individual.
I swap between Fable and GLM 5.X and both work fine. Fable tends to do stuff that I didn't want it to do, and claude is probably a bit more capable as a harness (though I haven't spent the effort configuring pi properly yet so that's probably on me). Otherwise it's hard to spot the difference.
I think you're off by a few orders of magnitude on cost.
People generally don't have 10% of their income to just give away on a whim.
Additionally, if I'm 10% more productive I don't get a 10% pay increase.
> You can move mountains with Astra or Opus 5.5.
Mountains that were made out of molehills, an ever-growing heap of bullshit.
The list of people who have made millions and would be indifferent when the source of those millions is being challenged can not be very long.
Really? Pulling the ladder up after making your millions is almost idiomatic for those people.
> It’s easy to be optimistic when you're sitting on millions (dhh).
It’s natural to be pessimistic if you are unskilled. Your ability to keep your job is directly proportional to how good you are.
Nobody believes that
How can Rails live long if no one creates new things with it? Our web applications aren't like the Pyramids at Giza. DHH literally said that even without a blank canvas (Hey's backend) they were still able to re-write the backend in Rust and are not looking back.
That sure sounds like any savvy business should pivot away from Rails and Ruby in general as fast as they can, assuming they have tokens to burn. Personally, I built my career on Python, but any new thing I build is almost always in Go or Rust. Why mess with an interpreter if you don't have to?
Having built two apps in the last two years using Rails (one making front page on Hacker News and has 10,000 plus users), there is a sweet spot that Rails hits that other frameworks don’t.
The fact that you get so much out of the box that just works and you don’t have to do much beyond configure it reduces a ton of headaches.
Absolutely would rather be using Go instead of Ruby but for apps of a certain kind and size, the omakase approach of Rails still wins out.
He clarified today, saying he would build any new app today in Rails. Should that app become so popular that it would benefit from Rust, he would rewrite it in Rust.
Source?
According to the schedule, today had a keynote on "AI and the future of Ruby & Rails — a chat with Matz & DHH". I suspect that might have clarified, but sadly they don't seem to have released a video (yet?)
This might just be damage control or politeness more than the truth.
I agree with him. All software should be written in safe languages, but there is a wide selection available. For new projects you should pick a language that does as much for you as possible. Ruby and Rails both go out of their way to provide the developer with as much power as they can. They make development less expensive. That’s exactly what you want when you are first building something, before you know exactly what you will need.
Later after you have lots of users and your hosting costs are rising, then you can rewrite it in a language that gets you closer to the machine. You won’t need to do any exploration at that point, and you’ll already know where the performance hot spots are. Of course I recommend Rust for that, but that is merely my personal preference.
Yes, true enough.
But there is a "dead middle".
If you have a big cash flow to go with "...lots of users and [rising] hosting costs" then, yes, you're spot on
But if there is insufficient cash flow then costs are rising and you are a bit stuck.
You can accuse DHH of many things, but not of grifting, insincerity, political politeness/correctness just not to offend people. The guy speaks his mind and is unapologetic about it.
It would have been the case if he said all the things in keynote unprepared like in a podcast. i.e Speaking his mind.
But it wasn't.
I often need to prepare to speak my mind
How is DHH not a grifter? Is the white replacement conspiracy theory he promotes suddenly not a hoax?
The drastic change in the ethnic/cultural makeup of people leaving in many Western countries is a fact. In Finland it is +70% foregners only in the past 10 years. Do you want to dispute stats? Write to the Finnish department of statistics.
Or are you arguing that this is a natural process and politicians can't influence it? Wrong. Pre-post Brexit stats show the opposite. As do the stats from Denmark, after politicians started working on the issue.
You are fine with stats, but don't like DHH's opinions about those facts? Well, those are opinions and everyone has a right to them. Opinions and feelings based on morals can't really be proved incorrect.
Or are you tone policing?
So what exactly is the actual problem? "Someone on the internet has wrong opinions and is not a carbon copy of me, politically"? What a silly and totalitarian behavior to call other people dirty names!
Stop pretending that racism and xenophoby are just political opinions. They are not.
Stop calling anyone with differing opinions a racist or xenophobe. It really feels like these activists are deliberately misusing powerful words in order to weaken them (if everyone is called a racist, no matter how undeservedly, then no one is). Or they are simply stupid and do not understand the reaction they are causing with their totalitarian tactics.
Racism is also an opinion some people choose to have. If people don't like being called racist there's a very easy solution to that.
Oh well that's... um.
> Why mess with an interpreter if you don't have to?
I mean I haven’t touched Ruby in years but if you’ve got tokens to burn, who cares? Compute is pretty cheap these days and Ruby is usually fast enough for web stuff, and as far as I can tell 2nd hand the robots write it just fine.
And I say this as someone building a strongly typed programming language that compiles to native code.
Maybe not the most popular answer, but... because JIT?
I actually have no problem with him even if he directly say it out loud announcing the death of Ruby and Rails. And I am not even against agents taking over some jobs meaning less jobs for current developers. That is part of how new technology has always been, replacing some and we move on.
It is just that it shouldn't be done in the opening keynote of a conference. I thought it was very irresponsible to the other smaller sponsors and those who paid to join. I know Shopify fund most of it and is likely on the same stand.
> In the meantime… keep showing up. Keep wondering.
> Long live Ruby. Long live Rails.
> p(bloom)
These aphorisms try to put a more positive spin on the keynote, but they miss the point that made it so disappointing for a lot of people.
They travelled to a Ruby on Rails conference because they wanted to hear RoR talks. The keynote was all about AI and Rust and ended with “don’t be losers” about the AI debate.
Ruby on Rails content not found.
I suppose the die hard fans who have been around this long and who continue to travel to see DHH speak have aligned themselves with him as a personality figure, maybe more so than with Rails itself. Outside of that bubble, it just feels sad that one person used the conference to do yet another controversial move in an inflammatory way while ignoring the topic everyone travelled for. I say this as someone who uses a lot of AI and also agrees that programming work has changed. It’s just so insulting to be surprise-lectured about it with condescending “don’t be losers” comments when the topic was supposed to be Rails.
There was an opportunity to talk about what AI meant for the future of Rails, but instead he went off bragging about rewriting Hey in Rust. Rust has been my primary language for years, but I don’t want to be lectured about it in a Rails keynote while Rails discussion is nowhere to be found.
I think his talk is all about what AI means to rails. It's dead.
Those who travelled to a soon to be dead web frameworks conference might appreciate the heads up.
Those who refuse to accept or believe it might be salty.
> Those who travelled to a soon to be dead web frameworks conference might appreciate the heads up
If it was me, I would have preferred to be informed it was dead before spending $699 on tickets and buying airfare and hotel reservations.
lol, yeah wtf. DHH's keynote could've been a blog post and saved everyone the money and time. But of course, we are in the Grift Era of tech.
I’ve made a similar point to my colleagues recently about cross-platform vs native apps. If we’re using AI anyway, do the arguments we made years ago still hold today?
I know some people might disagree, mainly because AI isn’t a new architecture or a new programming language in the traditional sense. But I still think it’s worth revisiting the assumptions and having the discussion.
Publishing mobile apps in Apple store or Google Play is a shit show.
You have to adhere to bizarre requirements, you can be taken down on a whim without possibility to appeal.
You have to pay for a developer account at least to Apple each year $100, Google one time fee $25.
Nothing like that for a web app.
Put aside the usual web-dev perspective for a moment. As developers, we build for our users.
The web and native apps serve different purposes. Users do not use a website the same way they use a mobile app, and the trade-offs are different.
I’m getting tired of web developers failing to acknowledge that and treating web technology like a hammer that needs to be applied to every nail they come across.
And especially when people say “the web”, what they often really mean is Chrome and the Chromium ecosystem. A lot of web apps barely get tested properly on Firefox or Safari, which only reinforces Google’s monopoly over the web platform.
Yes, app stores have their problems. That does not make the web the right answer for every product.
Put the money where the mouth is.
If I name the price for actually developing apps no one will pay, there are only nerds nagging about how they are annoyed by web apps.
I think more often the comparison is web techs (like React native) VS native apps.
It’s not really as bad as you make it sound.
You forgot the 30% revenue share the App Stores take.
For a lot of companies it is not a factor because they have free app complementary to their services.
Main point is that even if you could slop out some code to have a mobile app there is enough of additional burden that makes it not really worth it when you can get away with web version that is good enough and you don’t have to care about all that bs coming with publishing and maintaining a mobile app or to be specific you need to maintain exactly 2 apps.
It is that bad.
I mean, that’s like half an hour of tokens, right? (Half joking laugh/cry etc)
Depends a lot on the token budget I think, different situations for established businesses or strapped founders vs side hustles or hobbyists. So in a way you outgrow Rails or frameworks like it once you can afford a software factory on autopilot 24/7.
shopify agrees https://shopify.engineering/back-to-native
With cross-platform frameworks, you’re only paying once for tokens. As businesses’ AI costs continue to rise, that can be a huge decision factor.
DHH is right (pun intended), p(bloom) is the ONLY WAY to LIVE.
Living in the fear of p(doom) is miserable. We're all gonna die one day anyway.
That worldview assumes that technology is like rain or an asteroid shower, and we just have to accept it. Its not. Humans make technology and decide what is going to be used.
No technology is inevitable. Technology isn't nature.
The constant has always been tools that maximize developer productivity. What's changed is that LLMs are now the best way for developers to tell machines what to do. We've all been given a much bigger lever we can use to spend our waking hours (and even sleeping hours if you run agents overnight) to move even bigger mountains.
robbyrussell, creator of oh-my-zsh?
1. thank you for omz, it's great
2. how is it possible you've been on HN for over 15 years and have just 254 karma?
Lurkers abound.
How long will that last? It's evident that they are trying to squeeze the engineers to the max.
Ask the COBOL guys
The market being what it is, I took a job in ColdFusion last year.
There are people here who have worked in the same CF codebase for over twenty years. They might start and finish their careers in a language everyone else forgot about decades ago.
I think Ruby has enough momentum that career Rubyists will have a home for as long as they want one.
How'd you find the job? And was it remote?
I spent 2 years coding in ColdFusion maybe 25 years ago. I've got experience!
Smart. Bravo.
Buddy of mine pivoted to doing just SQL and road into retirement as a DBA.
With hindsight, we both wished we had gone all-in on SQL way back when.
That is actually a smart move. You only have to learn SQL once to use it for 50 years.
The foundations yes, the rest only if you never change between RDMS products.
On my 50s, learnt SQL around 1996, have used SQL across Access, Informix, DB2, Sybase SQL Server, MS SQL Server, Oracle, MySQL, Postgres, PrestoDB, Neo4j.
They’ll be getting nervous too at some point the agentic port will come for cobol. The past hesitation and risk aversion will subside once the less risky stuff is addressed. Once someone big enough sets a precedent, it will gain momentum
> the vibe here is far from doom and gloom.
To be fair, this is true of most of AFK world. Things start seeing a lot less doomy and gloomy when you're actually outside talking with other humans. Seems to mostly be online and various darker corners where they're dooming and glooming.
I get why DHH would be really happy about getting rid of engineers. It's less clear how this isn't a little bit gloomy for the engineers.
Is the idea that they'll switch to doing "product owner" roles overseeing agents? I don't see that employing the same number of people or paying as well as engineering used to.
To be honest I expect that that role to suit different people to those that used to be engineers.
If this feels gloomy to some people, it's because it is.
Why are you saying he got rid of engineers? Did you watch the talk?
Yes, I watched it. I'm not saying he got rid of engineers so much as just not hiring new ones.
If he's hiring new people they won't be engineers or the people that used to do engineering.
They'll be people that are better suited to writing English than code.
I don't think that hiring the engineers was ever an option. He said that he couldnt maintain six native apps before agentic coding. The business never even considered expanding headcount by 6X to do it, it was a complete nonstarter.
There is no fixed pie of work that AI is taking from us. In this circumstance, increased productivity caused the company to just make more stuff, not lay off people (or avoid planned hires).
It is a real world example of https://en.wikipedia.org/wiki/Jevons_paradox
To clarify, my point isn't that the market for software won't expand due to the greater productivity unlocked by AI.
It's that the benefits of that productivity won't go to the kinds of people that previously wanted to stay in engineering their whole career.
I'm not sure the Jevons Paradox applies because you're not just making engineering more efficient you're replacing it altogether with something much more commoditised.
People will gain from this but it won't be the same people.
Likewise, consumers will eventually reduce the amount they’re willing to pay for software so it will balance out. It will be a blip in time we look back on. Like, remember when elevators used to employ an operator. Now we just push a button and happily do it without another person’s labor.
Well at least in this particular case, it was the same engineers on the team. He specifically said he tried getting UX people to vibe code things and the results didnt work
Jevon's paradox has to do with consuming resources, not employment. There are plenty of examples of fields getting more efficient and employment falling. Farming and manufacturing being 2 examples.
You're conflating different categories of activities. Farming and manufacturing are amenable to automation precisely because not much real thinking and communication is required.
SE is not like farming.
It is a real world example of https://en.wikipedia.org/wiki/Jevons_paradox
It's an interesting example in that a switch from web/React -> native seems like Jevons paradox from the perspective a developer considering "amount of effort I'd need to put in"
But Jevon's paradox is about consumption of a resource increasing. The resource being "developer effort" but the new supply all running inside GPUs and dev machines. He actually says he decreased their server deployment by 90%.
So on net, did they basically deliver their users a much nicer but functionally equivalent app, which in total now consumes less resources / spends less $ into the economy? And then the $10T question is to what degree this is representative of AI's effect.
37signals typically doesn't do hire a lot of software engineers but just last week they hired Rob Zolkos who is very clearly a software engineer.
https://x.com/jasonfried/status/2100668586851537195
the algorithm is programmed to deliver it. Better engagement metrics with misery and anger than joy and hopefulness. (yes I know that you will reply 'but I prefer the joy and hope', but that's not the case across the wider population)
And I think the online reality is much closer to the truth than AFK reality. I can't think of a single founder I've talked to in the past two years that's building with rails. That doesn't mean rails dies overnight, but it does mean it's simply all downhill from here.
From your profile: CEO/Founder of Planet Argon...
Looking it up: Planet Argon is a Ruby on Rails development agency.
> a Ruby on Rails development agency.
He's sitting in the front row at a Ruby on Rails conference - what did you think his job was?
Probably not a lumberjack.
Promoting his rails services where the creator and biggest defender of rails replaced rails with an AI generated rust version without writing a single line of code.
His comment is just damage control and biased, it did not state he's heavily invested financially in rails. Not everyone knows who he is, I didn't.
I think most people on HN know robby russel. Have you ever used ohmyzsh?
Robby is well known in the Ruby community.
Concerns with his politics aside, I do think there's a truth to what he's talking about here, and that he is just spelling out the reality that developers are, or shortly will be, facing. For many this will be deeply uncomfortable to hear.
It is notable that his perspective in this talk is very much from a developer-user side rather than someone who is responsible for the framework itself. That surprises me, and I suspect it is not a good omen for Rails.
For all of this embrace of agentic development, there is nothing here on how they are adapting the framework for this new agentic development reality. Agents do currently work well with Rails, but there's nothing here pushing things forward as best as I can see.
I know for my own projects I've largely moved to Elixir/Phoenix, for similar reasons to his use of Rust... I didn't want to have to learn it, but now I don't have to and I get to benefit from its strengths.
I think a lot of people haven't actually watched his talk and don't understand what he is saying. If you watch the talk, he is not anti-rails. He is just very pro- agentic coding.
What he was saying in his talk was that his customers (people using Hey for email and calendar) never really wanted hybrid web-native apps, they want native apps (like most people prefer on mobile). His business could never afford to have a dev team that built and supported six different native apps, so they always built using hybrid web-native stuff using Turbo Native.
Now that his development team has access to agentic coding, they can build and maintain six mobile native apps. For his app, the back end part then becomes small and simple and they can reimplement it in rust to reduce server cost.
His point wasn't that rust is better than ruby, or anything like that. His point was that agentic coding allows smaller teams to do bigger things. For most people building on rails, their customers don't need native mobile apps (I know mine don't) and for them Rails is a fantastic choice for agentic coding.
Rails is so into agentic coding that they publish benchmarks of how effective different models are for building with rails:
https://rubyonrails.org/ai
TLDR: DHH isn't anti-rails, he is pro-agentic coding. Rails is great with agentic coding. AI doesn't mean layoffs, it means your teams can do much more than they could do before.
It seems to me, so far, everyone moved one level up.
- A layperson can now build one a solo developer can do
- A solo dev can now build what an agency of web development can do
- An agency of web development can now build native apps similar to what large corporation can do
- Large corporation can now tackle even more ambitious projects
Which makes sense given what got automated. This is with the current tech, people still exploring the full potential, but if that turns out to be the case then I'd argue the software market will grow substantially and we are just having an adjustment period.
This matches my experience. My solo projects are now bigger than I ever would have accepted alone. The one I’m going to finish soon is one I actually refused a year ago because it was too big.
It’s the top of the chain that I wonder about. Agentic development is so fast (and in the process I’ve developed so good) that I’m thinking there is actually a maximum size a company doing this work could get. Beyond that top end size the management overhead is what’s limiting or the market size total. Maybe not, maybe mega dev houses will find other problems to solve, but I have a hard time imagining it.
I do suspect there will be many mega projects. Things like new operating systems, new apps ecosystem (not just apple/android), new virtual reality environments, massive interconnected web of agents, in-house models etc, also alot of re-writes and update of existing infrastructure.. competitors to existing social networks and many established players..I think the entire software ecosystem will grow larger, wider and deeper from all directions.
From what I have seen the pie will get way larger, but there is growth pain and adjustment period for sure.
But to your point, I also think solo devs/freelancer are now seeing the most gains, that's because they can adapt the fastest and they don't need to deal with existing overheads.
Larger organizations might need more tooling/agents and better internal processes before they can see the full potential. I don't think ww, as an industry, are there yet, this tech started truly working less than a year ago.
> If you watch the talk, he is not anti-rails.
He explicitly said he still loves ruby as a language and thinks Rails is well suited for a future with AI coding.
It's interesting how this stuff gets spun on social media by people who don't watch the content.
It looks like he spent about 4 minutes out of the 1:03 making that point.
I can't but help think that he would have spent longer on it if that's what he really thought.
I don't think he believes in Rails any more.
It's clear that he was more excited about thing that are not ruby or rails and now considers rust better suited to the output of AI.
To be honest, I could see rust taking over from typescript/javascript as an output language for the web too. If no one needs to read source code any more why not use rust for everything?
I don't think anyone should really believe in Rails at this point. It was a monumental leap forward at the time but other stacks have shown that there are better ways.
>It's interesting how this stuff gets spun on social media by people who don't watch the content.
I watched it live, and a few actually in the show have very different perspective to what you suggest, they were just keep it to themselves and not posting it on twitter. Some did, but then deleted it.
And no one is questioning he doesn't like ruby anymore. But it was obvious there was nothing much said about the future of Ruby and Rails.
It doesn't matter if Rails is well suited for AI or not. Now, there's less incentive and reasons to use it.
When I started as rails dev in 2013, it was really the top solution for producing MVP fast. Now, I don’t think that’s the case and most of the incentives are gone if we are no longer writing code by hand
I prefer a Rails backend, especially in the agentic world.
With an agent on a greenfield is when I’m most reliant on a framework for good standards and a clear architecture.
This is one of the things Rails does really well.
So what would you pick for producing an MVP web app fast that you also want to run in production, maintain, patch and scale that has a solid eco-system of pre-built libraries you can just plug in to solve problems?
I strongly prefer Rails with agentic development. Rails comes with nearly everything you need to make a full-featured application, and that makes it easier and faster to use with agents.
No other framework comes close, and telling me that I can have my agents cobble together this same baseline, badly, from a hodgepodge of new code and random libraries is not compelling.
> AI doesn't mean layoffs, it means your teams can do much more than they could do before.
That sounds like layoffs lol.
At least within my circle, that’s not really the case. Instead, what I’m seeing is more of a hiring freeze, with the budget being used to empower existing employees through AI. Which I think is a more reasonable take.
That's often layoffs in disguise. If you want to cut 10%, but you don't want to do it as a single layoff, because that's expensive, can trigger certain laws depending where you are, and might disrupt teams, who do you cut?
Instead you can do layoff by attrition, you stop hiring, and you don't backfill seats, as people naturally leave or get pip-ed, the team shrink, you don't rehire into the position, and in a year or two, you cut down by 10%.
Also, I wouldn't trivialize a freeze in hiring growth, it means to get a job you have to wait for someone to retire, and also wait for people looking for jobs with more experience as they're more likely to fill the roll to also retire. Basically there's now way more demand for jobs than hiring, so that will drive compensation down for those currently employed.
Adding on to your point about attrition: if internal promotions in engineering are decided by your ability to be a force multiplier, to facilitate cross team collaboration, mentor and help others to skill up, how do any currently employed engineers demonstrate that with no one new to mentor or train? How do you collaborate when teammates work with AI tools more and coworkers less? Would they just have an easier time getting into a higher responsibility role looking elsewhere?
>hiring freeze, with the budget being used to empower existing employees through AI.
We can't hire or give anyone raises, but we will expect you to be more productive via throwing money into the slop void.
> We can't hire or give anyone raises, but we will expect you to be more productive via throwing money into the slop void.
i see your point and i want to tell you it's pretty much the same with cloud computing, which is so much more expensive than running your own hardware... and yet nobody's really questioning cloud computing (except the oxide people, but they're still tiny)
the company i work for could probably spent half its yearly cloud bill in hardware and be fine for 3-5 years.
guess who discovered the same thing and booted google cloud off their books? dhh
that guy is simply enlightened.
Even weird fascist broken clocks can be right every now and then.
it sounds like leverage
as a solo: you can build and ship more in your own time
as a startup: you can iterate and scale more with more velocity and less growing pains
as an enterprise: people with ideas but not a technical background can put their tokens where their mouth is and build their own interfaces, personal apps, ask their own questions, ship their own features without having to navigate layer 8 and play games for resources
Why does it sound like layoffs? If you keep the team size the same and use the productivity increase to deliver more business value, how is that layoffs?
If you can get more done with less, your choices are either do more by spending the same, or spend less and get the same output.
Why do you think 100% of all companies will choose the first? Obviously some will choose the latter, hence it sounds like layoffs.
Because the limiting factor for what you can deliver has never been, and still isn’t, developer productivity.
Talking point: I think it is communication, and AI (notably information exchange in “Claudish”) is not as effective there.
with more productive teams you can have smaller teams and better communication
Thus layoffs, no need to keep as many employees around.
There is no MBA like magical growth, the amount of customers is a physical limitation of inhabitants on planet Earth that are willing to actually use a specific piece of software, regardless of how productive it was made in first place.
If I figure out a way to make my workers produce 10 times more widgets each, it doesn't automatically follow that there is a market for the extra 9 widgets. Otherwise I would have hired 9 extra workers before that invention to fulfill the market demand.
Making the widget cheaper doesn't necessarily create demand either. I will not buy 10 times more toothbrushes or writing pens if their prices suddenly decreased to 1 cent.
The demand for software or features will not increase at the speed of our ability to produce them.
That is the fallacy of all those MBA degrees with exponential growth expectations, they never take into account the physical limitations of goods, like your very good examples.
It’s layoffs if the demand doesn’t grow to meet the output
If every engineer is 10-100x more productive with AI, and one company slashes its workforce to maintain the same level of productivity at reduced cost, it will lose out to its competitor who retains the same workforce and does far more with it.
Right now, the vibe at my company is not “we can save so much money with AI” it’s “here’s an infinite budget, use all the tokens you can to expand our moat because we’ve never been more sensitive to competitive disruption than we are now”
Competitor won't retain the same workforce if the demand doesn't match the output.
He's not anti-rails. He just won't be using it.
> His business could never afford to have a dev team that built and supported six different native apps
It’s his business decision to save on native app development costs to increase his (generous?) profit margins.
One of the things I admire most about Basecamp is that they are a bootstrapped, profitable business. There's no shame in running a business this way. The world would be more resilient if more business owners grew sustainably instead of soaking up endless invester funding and losing money year after year before enshittifying in a desperate attempt to repay their overlords.
Yes! I wish I had more than one upvote to give!
Here, I will give you my upvote! That's how Issuetrak is. The founder (me), still owns a majority of the stock, there's no outside investment, and we have customers who've been with us more than 20 years. Multiple employees as well!
>His business could never afford to have a dev team that built and supported six different native apps
I have a hard time believing that, considering his hobby of racing cars and commissioning one-off hypercars. Unless it's that his business could never afford that dev team after DHH takes his salary.
It’s not his message, it’s the forum. Why host a rails conference, that he explicitly created because he was silenced from the main one…to them not talk about rails?
Why not adapt rails for this next generation?
Why do it this way? It’s like walking into a temple to yell about Jesus. It was absurdly obnoxious.
It's almost as if, for him, it was more about him than Rails...
> TLDR: DHH isn't anti-rails, he is pro-agentic coding.
That's an opinion, not a summary though. The way how DHH behaved is really awkward in the last years. See the "let's shoot down wolves" blog entry here:
https://world.hey.com/dhh/wolves-sheep-and-gypsies-ba44af6a
That DHH is into AI agents everywhere is easy to see, but if you go to rails-con, would you not expect this to be about rails as such, and people? Now it's about rust - and AI. Controlled by big corporations. Yikes - some people must be upset about having gone there to listen to DHH preaching about an agentic future.
Edit: I also watched parts of the recent youtube video about rails. I could not focus on the content because DHH's style was so strange. I've seen that style before - nothingburger talks. Look at phrases he used such as "I have retired from being a professional programmer/software engineer". This is all aimed to hype up "agentic coding is the future". Rails is dead - that video is all the proof one needs. DHH shifted gears to AI, and it is quite similar to when he first promoted rails. But now it is about AI, not rails.
Do the teams fully understand the code that they are committing? Are the teams reviewing each others code or are agents? When agents do all the coding, and perhaps the reviewing, and the patching what is left of the team? People lose connection to their work, and the social bonds between workers begins to erode.(1) DHH's speech seems like a weird eulogy for the entire Rails community, although many of us saw this coming several years ago.
Also, I hate native apps. FWIW most devs I know actively try to avoid them.
(1) https://www.uclawjournal.org/how-ai-destroys-institutions/
Most devs I know who hate native apps are either web devs trying to embed web tech into everything they touch, or they just hate Apple.
Most users don’t care. What they do care about is performance, and native is usually the easiest way to get there.
For your AI workflow questions, it depends. With or without AI, those decisions depend on the organisation’s direction, the team’s capabilities, and their preferences. Did your org have a single answer for the best programming language, framework, or workflow even before AI? There was never a single right answer to any of that.
I hate native apps. There are use cases better suited to native apps but often it's just not what I want.
You have little to no control over privacy, tracking or advertising compared to the web. That's left to the original developer and there's nothing you can do about it.
You can't do things like disable autoplaying videos. Often you can't even copy text without taking a screen shot. On a phone it's also much easier to zoom random parts of the screen in a browser compared to a native app.
There are legitimate reasons to prefer the web to native apps.
Users absolutely care, people hate downloading random shit on their phone they don't want but are forced to.
What has always been true is that users often have no choice or voice to express their actual desires. It's not like there is a public alternative here (when there should be).
> Rails is so into agentic coding that they publish benchmarks of how effective different models are for building with rails
To each their own, but I think using Rails in the era of agents is probably a mistake. It has 2 big problems IMO.
(a) A ton of footguns that an agent can and will miss on occasion, and as we manually review less and less code, we will also miss it, and
(b) security on the web has become increasingly important due to hostile AI agents, and while rails has great secure conventions (see (a)), it's still easier to screw up something and not catch it until after it's deployed because there are simply less guardrails in place (dynamic types, no compile, slow tests, generally bad performance profile).
Rails dev here, still using it, still plan on using it, use agents, don't see them making food gun decisions with rails apis just code.
Rails comes with the console by default, so production inspection of a bug is something the ai can do without a ton of bs to be set up, and that console can be forced to be auditable (using standard gems) or read only (using flags).
I used to love rust, still do sorta, but yeah I think this is really not a good argument against a tool that operationally is better at most things at being observed or poked or prodded or tested in production
Not sure if you’re speaking from experience or just being theoretical here. We’re using Rails and it’s been great. Maintenance and upgrades have been a breeze.
All this static vs dynamic stuff doesn’t really matter anymore. AI doesn’t care. You just need to close the feedback loop with tests, linting, a compiler, or whatever else gives it a clear signal.
Simple as that.
If it doesn't matter I will just pick the cheapest to run (the static ones).
Yes, I'm still supporting two large-ish rails applications and a larger Rust application.
I don’t think this is agent specific. Rails is full of footguns that are virtually impossible to avoid even for a human developer.
What languages are best suited for development with agentic coding?
Rust is used a lot. Elixir/Phoenix comes up often. You’re looking for something statically typed with string guarantees and as FP as possible to make unit testing easier.
Rust for sure. Builds become a problem eventually though. Elixir is much more interesting now that they added types (I think they're doing more with this still?). I quite like Gleam from a human (non-agent) perspective.
I've been using Rust more and more for specific tasks in my Elixir saas app, the two play really well together.
In my experience, Go. It was designed to be idiot-proof. It's very verbose, not very expressive, compiles fast for quick feedback loops, strict autoformatter, plenty of examples in models' training data.
the only substantive difference I've noticed is between languages with a proof system and those without, where with a proof system is orders of magnitude better for agentic work.
OCaml.
Same. I find that the combination of the BEAM runtime and the guardrails that you get with an Elixir project, functional style and particularly the consistent structure of using Ash with it for data, API and CLI lead to a very maintainable codebase even with a lot of AI.
Phoenix with LiveView bypassing the need for a separate frontend framework is also a huge win there.
> Concerns with his politics aside, I do think there's a truth to what he's talking about here
One thing the AI era has changed for me is how easily I spot Gell-Mann amnesia, of which this is an unfortunately strong example.
His politics, even putting aside their nature, are framed with deliberate misuse of facts and data; why would he be someone to listen to in any depth on any other topic?
Gell-Mann amnesia assumes a journalist is not an expert on _anything_. In this case, DHH is (according to some) an expert on software but it's much less clear he's an expert on politics. So I'm not sure Gell-Mann amnesia applies here, it's maybe more like Nobel disease, where you're an expert on one thing and assume it "transfers" to other fields.
> Gell-Mann amnesia assumes a journalist is not an expert on _anything_.
Incidentally, this is also the fundamental flaw with Crichton's original description of the effect which questioned the accuracy of reporting about international affairs. When it comes to politics, simply having access to important leaders and meetings and reporting on them accurately and fairly is the only "journalistic expertise" you need.
Of course, the even broader problem with the concept is that, in reality, there are some articles written by journalists that contain a reasonable understanding of the subject matter, and some articles that do not. So seeing an article written about your field of expertise that contains misunderstandings doesn't by itself tell you anything about the distribution of good and bad articles covering any particular field.
> much less clear
I have a background in sociology and urban planning. It's very clear that DHH is not an expert on politics. His blog posts demonstrate a very elementary understanding of the topics he covers. He doesn't even understand statistics.
> I have a background in sociology and urban planning
What an interesting way to make it sound like that means anything. "Background" - how appropriately vague...
And how exactly does this make you an expert on European, and especially Danish, immigration, culture and vox populi?
People say "background" here, because the default assumption is HN readers have software expertise, but it's frequently unclear what else, if anything, they have here.
I do agree "background" is vague, but I generally assume the minimal threshold to say that is something like a previous career, undergraduate degree, or graduate degree in the background. I personally wouldn't say I had a background in something I was only an amateur enthusiast of.
---
That being said, what on earth gives you the idea a substantial background in sociology WOULDN'T be relevant here?
Because I understand what sociology is and isn't, since I too have "a background" in sociology and urban development. As well as statistics, history and even futurology.
Appeal to authority is a bias for a reason. In this case it wasn't even that well executed.
A "background in software development" (an HR or a marketing girl in the website design sweatshop) does not give enough relevance and authority to speak about the technical details of Linux kernel. Vice versa.
>His blog posts demonstrate a very elementary understanding of the topics he covers. He doesn't even understand statistics.
Examples of these claims?
Ha. I was trying to be charitable there :)
My Gell-Mann moment was watching other HN users rate Gell-Mann moments while not realizing they're demonstrating it directly.
That evokes the lesser known corollary:
Crichton's Law: unwittingly serving as the best example of one's own snarky law.
> His politics [...] are framed with deliberate misuse of facts and data;
Any specific examples? I am/was out of the loop on this and had to Google "DHH politics" to know what people are talking about.
With his infamous "As I remember London" article [1] he directly correlated skin colour to nationality leading to the claim that native Brits had dropped from 60% to a third.
If you check now using current sources the British-born population is still around 59% the only thing that has changed is that fewer people identify themselves on the census as 'White British' which has a host of reasons around it. So he took a misleading statistic, interpreted it incorrectly and then built some kind of nostalgic story around it to support what was essentially an excuse to support far-right thugs protesting.
[1] https://world.hey.com/dhh/as-i-remember-london-e7d38e64
The first link in his article cites his source:
https://en.wikipedia.org/wiki/Ethnic_groups_in_London
>So he took a misleading statistic, interpreted it incorrectly and then built some kind of nostalgic story around it to support what was essentially an excuse to support far-right thugs protesting.
His exact words from the blog:
"In 2000, more than sixty percent of the city were native Brits. By 2024, that had dropped to about a third."
White Brits are the only native/indigenous Brits to London. It's not misleading, it's just a fact. He did not interpret this fact incorrectly.
Native means that your place of birth is in a country and that definitely doesn't apply just to people who cite their ethnicity as white British.
Happy to hear I'm a Native American.
That's disingenuous because it's a neologism based on the population that were in America at the time of its colonisation. Native literally has the etymology of place of birth in the word.
Oh, well then I'm a native American, I reckon. The other kind.
DHH was talking about the Native British. Not native British.
No, this Americanism is factually wrong.
You're mixing nationality, a piece of paper, with being "native", an ethnic group. Does not work in countries across Europe, Japan, Korea, etc. The settling of the USA and its cultural dominance has broken peoples brains. Europe is inhabited by its indigenous people!
The English are an ethnic group (ask 23andMe) and they are the founding stock of Great Britain. Source - the far right extremist wikipedia: https://en.wikipedia.org/wiki/English_people
London was 97.7% white in 1960. Now it is around 50%, and falling. Again, source: https://en.wikipedia.org/wiki/Ethnic_groups_in_London
Is that good or bad? Leave that one up for debate, I mostly care about facts.
> London was 97.7% white in 1960.
I don’t actually believe this statistic, as an aside, but if we stipulate that it is true, it still doesn’t remotely mean 97.7% “ethnic English”. That’s a radically wrong implication.
It’s a little closer to DHH’s point about the percentage of British people (the Welsh, Scots and Northern Irish are not simply “ethnic English”).
But it’s still very wrong. At that point it was a ninteeen-hundred-year old working international port city, it has had a large population of Jews of European descent for centuries (the first large community being so notable that they were discriminated against in Magna Carta, after all), various displaced populations after WWI and WWII, a very large Irish population, thousands of Italians, many Dutch.
In the 1800s French Huguenot refugees and their immediate descendants made up perhaps ten per cent of the population of London (figures vary, considerably, could be 3% or nearer 15% depending on the boundary definition you choose and that at the time of their arrival the UK had no formal immigration policy so no exact numbers — but over 200,000 settled in mainland Britain). A truly astonishing number arrived and the city simply made room. But they didn’t leave; they assimilated. Same thing happened with Dutch exile craftsmen.
This happens a lot in London’s history: https://refugeeweek.org/wp-content/uploads/2015/10/Historyof...
So he is wrong (and IMO ought to have known some of this) and you are not at all on the right track either.
People just have no grasp of the nearly unique reality of London as a historical melting pot. By 1960 its diversity was eclipsed by American cities like New York, Chicago and Philadelphia, but this idea that London was somehow traditionally “ethnically British” to the point of being a quaint monoculture is fully crackers. Just not true. The difference since 1960 is just more visible because we started adding more literal colour from the Commonwealth.
A place worth visiting for Brits who don't believe this and tourists who are interested:
https://www.londonmuseum.org.uk/docklands/
Super-cool museum space dedicated to London as a port city.
The literal definition and etymology of 'native' relates to birth, check a dictionary and the definition will read the same. The truth is that 60% of London's population is made up of native Brits and that hasn't changed at all over the last 25 years.
this is so tiring. you obviously know what he meant by native. the way you painted him looked like he made obvious errors in statistics.
you obviously know this because you know what "native americans" means.
Claim to be a Native American then.
https://news.ycombinator.com/item?id=49839539
> White Brits are the only native/indigenous Brits to London
Bullshit. They are Johann Come Later Germans who killed off the Celtic Brits.
You think Celtic Brits don't count as white? Have you seen us? We're some of the pastiest looking people you will ever meet. I could get sunburnt opening the fridge.
A friend of mine phrases it as getting sunburn from browsing Expedia.
That is also bullshit (although a popular myth, deliberately cultivated by 19th-century English supremacists to justify subjugating their Celtic neighbors). Every region in Britain has a majority of Insular Celtic DNA on average (though the fraction varies). Even in the most heavily Anglo-Saxon regions like East Anglia, Germanic DNA is in the minority.
https://www.ox.ac.uk/news/2015-03-19-who-do-you-think-you-re...
And of course the Celtic Britons replaced the Neolithic farmers, who replaced the Mesolithic foragers, and so on. "Indigenous" becomes meaningless with so many population replacements.
The fact he used the term “native Britains” says everything about his politics.
A proper British nativist would go full Tolkien and start resenting French influence on the language.
Thank you for mentioning the effect by name. I read first about it 20 years ago, memorized it under a totally different name, and was never able to find it again.
I am slightly abusing/overstretching the concept, to be fair; other commenters are not wrong to remark as such. But still.
One thing the AI era has changed for me is how easily I spot Gell-Mann amnesia, of which this is an unfortunately strong example.
The one for me is spotting the Baader-Meinhof effect, of which this is an unfortunately strong example.
Touché. I mean it has almost the same meter. I get that you're disagreeing with me, but have my compensatory upvote.
> The Gell-Mann amnesia effect is a cognitive bias where a person spots major errors in a media report on a topic they know well, yet completely forgets that unreliability and trusts the rest of the same publication on unfamiliar topics.
Yes. I am broadening the definition of "media" and "publication", maybe, though it is posts on his own publishing platform and performances at his own breakaway conference.
But if someone misrepresents deliberately on a topic of polarising importance, why would you spend any time listening to them on any other topic?
I would (still) listen to him if I did not know he wrote stupid wrong things on other topics in bad faith.
Do you automatically dismiss everything someone says based on their opinions on other topics not relevant to their main job? Is everything always absolutes?
DHH or anyone else spouting opinions on social media is hardly qualified to be setting policies for countries. Social media is also hardly the place to have any sort of debate on those topics.
Aha, the common HN tactic of expanding someone’s argument until it appears ridiculous and then attacking that. But I will bite.
It tends to only be when someone is egregiously wrong or (as in the case IMO) acting in deliberate bad faith on a topic of massive public concern, like this. Because these people are a waste of psychological resources.
I totally write off proselytising antivaxxers who I think should have reason to know better, I tend to write off climate change deniers who should know better, and yes, I tend to write off educated people who mislead with statistics to create intellectual cover for xenophobic, bigoted thugs and their antisocial campaigns.
This is not an exhaustive list, but it’s a good one. Other people who get written off tend to have a nasty streak on a topic of some divisiveness. For example I am not religious but not anti-religious; I am anti being deliberately or loudly cruel, posturing or nasty about it either way. People who are like that can be like that, but they will not get to tell me about their professional interests as if their awful behaviour is just a quirk.
YMMV, but I am now middle aged and I don’t have the time for this anymore. Freedom of speech isn’t without consequences for loss of respect. If you think my position awful or lily-livered you are fully within your rights to let it colour your judgement of my opinions on technology.
Apparently somewhere between a third to a half of American scientists believe in some kind of god. Not to denegrate those who believe, but that's some kind of compartmentalization going on. Politics isn't any different. I dunno, I guess I don't have much issue listening to people with whom I disagree. Some people just get stuck on things. I do to, and maybe I'm the one that's wrong.
It’s not the same, though. I just addressed this in a reply to a sibling comment.
> That surprises me, and I suspect it is not a good omen for Rails.
I had a fairly opposite reaction: Knowing what you can't be and how much less important your past work is going to be is good. If there are good reasons for more people doing less in Rails, then the creator is probably the person you want to be most level headed about that.
(They can, of course, still be wrong)
By "concerns" did you mean to say "I don't agree with some of his political opinions".
I literally just patched an AI-generated API endpoint on a new service at work that shipped without any auth whatsoever, because AI was re-implementing the auth token check method individually in each child controller instead of implementing once in a before_action hook. That's Rails 101 stuff and the app is small. It was so obvious that I saw it right away just reading the code, I didn't even set an agent loose to do an initial inspection. It was a real "yeah I still got it" moment for me.
I'm on board with the idea that agents are going to write most of the code, but not checking it is just insane to me, based on some of the things I've seen committed in commercial codebases recently.
One of the strengths of Rails was that it helped developers who maybe didn't always know the "best way" to build things to build in a standards compliant way – not that you couldn't do things like return a 200 for Not Found if you really wanted to etc.
I'm always pleased when working with a dev who has Rails experience even if they've since moved stack, as often they have picked up and learned the value of these good behaviours. Perhaps it's as much a community thing as a framework thing, or combination of both.
Agentic development definitely still has room to improve, but I suspect that there is some future framework (maybe even language) designed to help minimise these types of mistakes.
Interesting that all the patterns Rails adopted for humans to write better code can't easily be funneled into the LLM doing the job now. Seems like it should be.
Is it types? Are types the answer? Is it rust?
I think the issue with Rails, as a long time developer in that ecosystem, is that much of the default assumptions are a) not common (not so far off-piste that a human can't get used to them, but outside the center of the bell curve), and b) not explicit. So it's very common to do things like install a plugin that has across-the-repo effects but that is only documented in e.g. the config directory and the bundle. Unlike, say, NPM, when you install a new package where you explicitly have to e.g. import it everywhere you want to use it.
As time has gone on Rails has gotten better about this, but it still means that "your rails" and "my rails" are different enough that AI could easily get confused.
As you say, the other aspect is the intense dynamism which works against AI, which again prefers to have everything explicit and reasonable about from the text.
So I ended up going back to Javascript after a brief 12 year digression into Ruby. I still love Ruby but it's just not the right language for the current environment and team structures - typescript is "worse is better" in the right ways that it's a much better choice for both frontend/backend being the same and better type systems.
Sorry, but what’s “install a plugin” in rails? That’s not a thing; not even a thing someone who’s ever worked on a rails app would say by mistake.
How certain are you about that? https://guides.rubyonrails.org/v2.3/plugins.html
Back in the very early days you would have used script/plugin install (svn url)
Here's one of my mirror repositories migrated from SVN into git. https://github.com/jaggederest/acts_as_voteable
Types are great. No nulls are great. Exhaustive switches are great. Derive macros and Serde are great. Detailed errors are great.
However Rust has one serious problem, BUILDS. It’s slow and it takes up a ton of space and memory. You will be working around this on any sizeable codebase. Using cargo check to ration your compiles, a dedicated build box, etc. Now imagine you're using it for a web application with compiled maud templates... you're potentially waiting minutes to see your changes.
If we accept that over half of the people building with rust are using agents (I'm sure this is much higher in reality), then you should also accept that there is a ton of time and resources spent on building / compiling. Optimizing this should probably be a top priority. They need someone who will be fanatical about shaving time and memory use from the build process.
Sounds like OCaml has the answer then. It has the equivalent of all the great things you described in Rust, plus compile speeds similar to or better than Go. Agents should be able to blaze through changes with it.
Sounds like I should check in on rust in a year or so.
Yeah. I'm going through an exercise right now with a large codebase where I'm split some of it up into crates so that iterating on specific parts incurs less build overhead. Literally changing my architecture... which annoyed me a bit, but it does make sense.
> Agentic development definitely still has room to improve, but I suspect that there is some future framework (maybe even language) designed to help minimise these types of mistakes.
Maybe there will be one day, but what are the odds that it will meaningfully move the needle for existing platforms such that manual code review is unnecessary? "Blindly trust the output" is not a feasible security posture today.
I still review PRs, but rarely suggest changes. The most meaningful reviews come from our review bots. I mostly review broad architectural decisions as a way to keep abreast of changes in the codebase. There's a cohort of engineers I work with who I would be perfectly okay with letting the clankers review, approve, and merge their PRs. But there's a larger cohort of engineers who need what I would call a directional code review.
Why do you think an AI audit (say from Fable/Astra at xhigh) wouldn't have caught this?
Software engineering is so back.
When designing systems, you want the important details to be right. Especially with authentication and authorization.
From an architecture level, you can know which classes are important to review and which ones are not.
Yeah that is how I see it.
In Software Engineering, we learned about requirements, testings, system design, UMLs, etc, and those seems to be more relevant than ever.
Agile is dead. It presumed writing the code was the slowest part of the cycle. Now it makes more sense to only start writing code when the requirements are known as the code is quick and low cost to change if/when the requirements change later.
There are two problems that arise from cheap code:
1. You get complexity back at the speed that you add code. Also the complexity compounds.
2. As a human you are still responsible for your code, and you need to understand how the important parts of the code base work.
When you combine these problems, you get a complicated code base where you don't understand the important parts.
That is why software design is getting a comeback. It actually helps with both problems. Less complexity and more understanding.
What model though?
But he himself promotes such a future too. This is a conflict of interest.
You can say "ride the wave", but this is more - this is helping the wave become huge. And cashing in too. Sorry, but I no longer buy any of what these guys do. To me they seem like BigBro ad-clowns.
> Agents do currently work well with Rails
According to whom? And not everyone wants to become dependent on corporate-agents sucking them dry financially. We all know the increase-in-cost will come with regards to tokens, after the get-hook addiction phase is over.
Or you know just use one of the many great open models. The LLM wave can be taken just fine without "corporate-agents suck them dry".
Concerns with the dogshit I stepped in aside, I had a wonderful time going barefoot through the tulips.
Nothing that he said is just him espousing the truth. It is the world as he and all the other AI-pilled people wish it to be. It is a world they want to convince everyone is inevitable. There are still a lot of people outside of tech bubbles that don't like the direction AI is going and don't want to embrace what these companies are doing or how the government is handling it.
And honestly, do people really think everyone is just going to become an "energized creator"?
AI is not as big of a productivity boost as compilers were but it is not zero. Back then, SQL was created and marketed as something “managers could use”, another parallel to the current cycle. This whole hype cycle is something we’ve seen before and IMO, we should know better how to handle it this time around. It’s a bit disappointing to hear the same arguments every time.
This has nothing to do with programming or productivity. I am addressing the fact that people like DHH and his fans are in the habit of framing the future as written and that you can either embrace it or get crushed. Embracing it is giving OpenAI, Anthropic, et. al your approval to keep on going. I'm not denying that AI can do the job a lot faster.
Keeping the comparison to compilers, people in the 60s refusing the transition weren’t crushed, their skills were devalued. The open models work well enough at coding to keep up without buying OpenAI or Anthropic products.
People who are expert at assembly are still extremely valuable.
70 years after the peak of assembly, that value probably comes more from supply than demand.
But as a percentage, there are a lot fewer of us then there used to be.
Given that we know he's literally a fascist (it's in all his political posts) that jives. "Embrace what we want or get mercilessly crushed" is a very fascist position.
i have nothing else to say to you other than good luck
Nah I'm good thanks. Call me old school, but I'd rather understand how things work, and try and keep a maintainable codebase. Rather than just get some random rust I've no idea what it is, or why it does it, pushed into prod.
I think this is going to be todays 'cheap outsource' culture. Good enough to keep shareholders happy, an apparent boost in productivity, before tanking the reliability of the company.
Remember when we used to scorn stackoverflow developers, or PRs that were more than a few lines long. An every morning git pull is now bigger than my terminal buffer, its complete chaos. Numbers game to keep the new metrics world gamification going. Its dire.
I want to agree with you, at heart I do, but:
1. AI is exposing the harsh truth that 99% of the code out there has always been throwaway code. Yet good developers, void of any engineering skill, treated it like the 1% that wasn't and called their hygiene and practices engineering.
2. Way too many developers that have made their careers or internet fame out of their technical proficiency understand that the shift is unavoidable. See Antirez implementation of Arrays in Redis or his Dwarfstar project. Virtually every line of these projects has been AI authored. And yet they are terrific projects with extremely strong architecture. And not a single line was edited in vi (his favorite editor).
> Virtually every line of these projects has been AI authored
There's a difference between everything being vibe-coded where you don't look at the code and don't understand anything and using your agent like a very fast pair programmer.
It looks like Antirez did the latter. From https://antirez.com/news/164:
> In the first month I just wrote the specification document...Thanks to AI, the specification evolved a lot, via back and forth of feedback, intellectual challenges about what was the best design, what was the right compromise, what was too engineered and what not.
> You know what was the biggest realization of all that? For high quality system programming tasks you have to still be fully involved
Judgement will be required even moreso now. There will be critical parts of the system where it's important that engineers know them well. But there's a lot of work around the core that isn't that critical to know in detail. For these types of things, I've found it's just important to understand the gist to make sure things aren't going off the rails. But for the core pieces, it's important to understand the details of what's happening.
I think you are both correct.
I agree that we have to keep understanding our code and we have to understand the systems we build.
I think the deeper issue is that AI can generate so much code that reading it simply doesn't scale.
Even if AI were to produce perfectly written code that didn't have any bugs that human review needed to catch, it would still be too much code to read in order to understand everything. The machine can write much faster than I can read.
Extending this argument further, it also becomes impossible to onboard young people to codebases if we maintain an expectation that reading all the lower level source code is how they come up to speed. I don't think this was ever an actual expectation in the industry, although we may have short-handed it that way in conversation.
So IMHO we need to come up with higher level visualizations and other ways to compress the knowledge that we used to get out of reading source code. We basically need to create new higher level projections off of our semantic understanding of the project.
A lot can (and should) be said about this keynote, but I have a hangup on portraiture as an example of "creative destruction" (how I loathe that expression!). Classical portraiture is still alive and well and artists are still able to make a living from it - possibly even more of them now than during the 1800s.
The advent of the camera did push visual art in new directions, but not quite as presented here. Oil painting and portraiture wasn't as static as it's made out to be, and there's a difference between a commissioned portrait and artists painting people because they enjoy painting people. Giuseppe Arcimboldo, El Greco and Hieronymus Bosch are some of the most well-known experimenters in western art and they were active in the 1500s. Some of the most famous art movements (impressionism, expressionism, arts & crafts) appeared well before photography was democratized.
The biggest difference, though - as much as I love art - is that portraits are a distraction. Then as now, it's mostly a vanity project for the affluent and of little consequence to the lives of the vast majority. Software, on the other hand, is not only a multi-billion dollar industry employing millions of people, it's also a direct matter of life and death, since it's an essential component in things like nuclear power plants, airplanes and medical equipment.
Hieronymus Bosch is such an interesting case, and I like your point in comparing oil painting to photography. Unfortunately I also share your concern about software's criticality. Therac-25 comes to mind as a case study taught to us in computer science university courses, that have haunted me in recent times: https://en.wikipedia.org/wiki/Therac-25?wprov=sfla1
> "creative destruction" (how I loathe that expression!)
It is sadly a fundamental feature of capitalism. We will see this again and again as long as our economic system is set up the way it current is. Just imagine, there will be a time in the perhaps not-too-distant future when the old folks of that time will lament how easy the kids of today have it!
So the keynote mentions that in a perspective you are going to be a "maker of things" rather than a coder. But, nobody is asking, in a perspective, why would anybody use the stuff you "make" rather than using AI directly? All of the notion of apps disappears, in a perspective.
Do you change the oil on your car yourself? No? Why not? Buying the oil, filters and all the tools you need to do it forever would cost less than one oil change at the dealership. Plenty of videos on YouTube you can learn from.
Sure, but changing the oil means buying a bunch of stuff and getting dirty. Replacing your vibe coded garbage app with my vibe coded garbage app is much much easier than changing my oil.
That said, I think this only applies to apps that can be replaced by vibe coded garbage. Anything with a network effect, or anything that requires deep domain knowledge is much safer
> Replacing your vibe coded garbage app with my vibe coded garbage app is much much easier than changing my oil
I am not sure what is it about the HN crowd that makes them so out of touch with what the average Joe is capable of. It's like that Dropbox post top comment all over again.
Don't you think that the only reason you think it's simple is because you're... A professional Software Engineer??
I know for a fact that it has gotten simpler. Which means that the number of people who can do a version of the thing i’ve been doing most of my life got way way way bigger in the last couple years.
Sure, not everybody can do it - i take your point. But the dynamic between can and can’t has changed. Projects I wouldn’t take on because they were too much work for not enough reward have become something I can pop off in a couple hours. People who know nothing about coding but a lot about their specific need can build complex apps in days.
> Which means that the number of people who can do a version of the thing i’ve been doing most of my life got way way way bigger in the last couple years.
Which invariably means that businesses who incorrectly believe that their vibe coded solution will withstand the trials and tribulations of production will be needing even more professional help when it all goes wrong.
Everything old is new again.
Yes people shouldn't discuss until they have tried vibe coding as SaaS. I do this now and it is 10x faster at coding and 1x on any other tasks. So overall maybe 3 months work is done in a month. This means the typical SaaS may be cheaper now but it is not like anyone can vibe code it cheaper than the subscription fee.
> Buying the oil, filters and all the tools you need to do it forever would cost less than one oil change at the dealership.
Are you changing your own oil? The cost of oil alone usually ends up being around 3 DIY changes vs one at the dealer ship. The mobile 1 synthetic I like is currently $92 for a 3-pack of 5qt containers. Which is enough for 3 oil changes with some left over to almost do a 4th. Dealership currently wants $84+tax for change w/free tire rotation. It's cheaper but certainly not "do it forever", and living in an apartment often means risking getting a warning from the landlord for doing auto maintenance in the parking lot since its against the stated rules in the lease.
> Why not?
Drain plug is in a stupid location that I don't often feel like dealing with. I do change the oil myself in all my other machines with much more pleasant for oil change designs, though. So I guess what you are saying is that AI isn't ergonomic enough yet, but will that hold true much longer?
I have been thinking quite a bit about this lately. I am in the process of building my own music software applications. This is actually allowing me to leave MacOS since Ableton was previously the anchor for me. I can now build tools that are more geared towards what I always wanted in Rust much like DHH was talking about in that talk.
I could sell some of my tools I built but I am on the fence about it. As a few others have said, there is still work and taste that goes into those tools that some might not want to put the effort into. However, there will be an increasing number of people who just build bespoke tools for themselves. That's probably ok. There's a market for both.
The most important thing now (and maybe always) is distribution. If anyone can build something then it ultimately comes down to your ability to market, build trust and provide support at 2:30 AM. That's hard to do when you're a one person shop. Software is an ongoing operational commitment if you make it your business.
Because it does take some skills, judgement, taste, knowledge, time and effort to build things. He said it is not very different from directing a team of contractors, and also said that a 10x dev in now 1000x with a tool like this. He also gave the parallel of photography to address this point, that people were forced to be more creative and adapt.
But I assume you didn't use to build anything substantial.
> it does take some skills, judgement, taste, knowledge, time and effort to build things
As of now, but we are talking in a perspective.
People are not well organised.
Business people I work with tend to have all kind of ideas in different places, I cannot count all the Trello boards they have abandoned. Users are usually lazy and don't want to fill in data properly, well yeah chatbox with llm is cool interface, but you still have to enforce structure for data so it is useful.
With AI problem is still there "garbage in, garbage out", models are not going to imagine correct data to fill in blanks, in the end if you need someones birthday, someone has to type it in.
Web app I am working on is organised only because we get engineers and quality assurance and we get paid for maintaining application. Data is structured and we can deliver insights because the application enforces structure.
There are far more requirements than AI can deliver on its own.
>and also said that a 10x dev in now 1000x with a tool like this
Yeah, I'm sure you are doing ~4 years worth of work per day. Get real Jesus Christ.
At least at the current level of AI, it still takes days/weeks/months to make a good app if you have standards. That's not effort everyone will want to do.
If/when agents become reliable enough, there will be no need for web apps, nor phone apps, etc. In fact, DHH got that part right: "your app should have a CLI if you want me to use it." If "English is the best programming language", then English is the best UX. I'm not navigating a silly page and filling out forms and dismissing popups if my agent can book a flight for me with a sentence and a bit of textual back and forth.
I wonder if the web will go away and the Internet will be just bot to bot traffic.
Navigating the web seems archaic in that context. Not only is that the end of Rails, but of frontend, JavaScript, etc. in the long term.
Geocities made everyone a webmaster, too. Remember those badass websites?
So far a Rust rewrite's success rate directly, probably super linearly, correlates with the presence of a good, well designed test suite. This is why Bun's rewrite and others are successful. I anticipate that this will be true for the 37Signals rewrites.
I want to learn a couple of things once they are done:
1. Once the test suite assistance is over, how do features get built in their backends and continue to maintain the same quality. LLM test writing is a bit trickier as it often favors what Randy Coulman called tautological tests back in the day. It takes a lot of work to get them not to do this, so I am guessing folks at 37Signals will read at least this part. I can't imagine zero Rust read at all.
2. Over the years they had a few innovations in their domain design, ones that Rails made easier to do, for instance their delegated type pattern. Will Rust replicate this, and how would one know without reading it, or perhaps it does not matter? Furthermore, how do newer patterns emerge? and how does our arsenal of better abstractions keep growing?
I don't have the answers but I am glad there is a chance we can learn said answers.
Classic DHH.Allways talks about what is good for his situation, not what is good for the community in general. I can understand his excitement of slashing his server and labor expenses. He will be richer and have more time, less people will work for him and wow, rails came a long way, didn't it ? The world is your oyster now! Oh and btw, we are changing the name of the conference, it's called Rust conf now.
That's kind of the point of conference talks.
"Here's something that I did that turned out to be helpful to my situation. Maybe you can learn something useful from it for your situation"
Mediocre talks maybe
It is quite telling how virtual no conference talks mention concepts like solidarity or technology that is useful for society as a whole (not tech that just enriches the few).
"This is the most confusing funeral I’ve ever experienced" is the top comment there. Rather apt!
Regarding the native vs web app problem, this was a completely made up human-made problem. I think I would have preferred a deterministic solution like a game engine but having agents code up native apps based on your web spec is not a horrible idea.
The bigger question to me, is the original implementation, are agents going to do that, or will humans hand code it. My two cents, I’ve been using agents heavily now, I review thousands of lines of code a day, I don’t think this works unless I have a clear idea of the spec I want to build in my mind already.
I get a better understanding of what the architecture should be and how I should build it by actually doing the thing, which means typing code at human speeds. I don’t reason my way into a good architecture, I don’t think I ever did that much in the past and I definitely don’t get it by asking the agent to check its plan and explain it to me the hundredth time. And then once I know the architecture, I don’t think the agent even speeds me up that much. It changes 7 days of coding to 1, but those 7 days weren’t worth much and I generate even more ideas and mental models in those 7 days that pay dividends for years to come.
Would you have told a tailor that using the industrial sewing machine improves his skill at making delicate, beautiful clothes? Of course not, actually even in 2026, large swathes of high quality delicate clothes are hand sewn. There will be an equivalent for coding, even DHH admits it, when he hedges with “most”.
I think, for me I’ll navigate my career to focus more on extremely high quality, delicate custom code and leave this agentic coding to the sweatshop workers. I wanted to do this anyway before agents, now I need to accelerate it. yes FAANG is sweatshop work.
I was hoping there will be news on Rails 9.0. Feels like this is the end of Ruby Rails.
Hey is not even a Ruby Rails App anymore. And indirectly admitting Ruby is slow considering they manage to cut 99.9% of CPU with their rewrite in Rust.
I hadn't connected the dots until you mentioned it, but yeah this keynote indirectly says it's the end of Ruby and Rails. Ruby was designed to be coded by humans, and DHH says it's the end of coding by hand.
Even the concept of "framework" as we knew it becomes different in this age. If you only want reusability, why not use plain good old libraries and let the AI assemble them as needed for your project?
> Hey is not even a Ruby Rails App anymore. And indirectly admitting Ruby is slow considering they manage to cut 99.9% of CPU with their rewrite in Rust.
Let's see how that goes.
Personally, I wouldn't want all my emails stored on a service where the dude doesn't even look at the code anymore.
Personally I wouldn't want all my emails stored on a provider that doesn't speak IMAP.
I don't think there's a non IMAP email service in existence. Even the king of incompatibility, Office 365, does IMAP.
https://www.hey.com/faqs/#can-i-check-my-hey-email-with-my-e...
https://www.hey.com/faqs/#can-i-use-hey-to-check-my-other-em...
In a sane world such a "service" would be outlawed.
Those performance numbers mean nothing considering they went from rendering every single view of a highly interactive app solely on their servers to moving all rendering to the clients. It’s possible the backend logic of the app could have run on 10 hosts running Ruby as well.
I agree it seems like the end of Ruby on Rails. I personally hope a fork will come along as a spiritual successor, because I still love building web apps this way.
Isn’t a lot of Rails dev done by Shopify and Intercom these days instead of by DHH companies?
And Shopify has been sponsoring Ruby JIT developments, exactly for the same reasons Facebook created their own PHP fork with a JIT, and how Twitter moved into the JVM after so many fail whales.
DHH is on the board of Shopify
Everyone knows that Ruby is slow, that is why there have been so many projects to add JITs to it.
EDIT: wrote Rust by mistake.
Ruby?
Thanks, fixed.
Agents might pick up Topcoat (inspired by Rails) as the framework for those Rust apps.
it is sorta sad that framework development will dry up with fewer incentives.
Unfortunately yes because Rails has always been a framework of extraction. Now they are not even coding anymore there is nothing to extract. No one is a programmer but everyone is a maker.
I actually believe new frameworks and languages will pop up that are explicitly made to supercharge agentic engineering workflows. We're still super duper extra early! November last year was the START of all this. We're still in September! Not even one year has passed!
Where will we be in 5 years?
Examples:
https://dioxuslabs.com/
https://lean-lang.org/
It doesn't make much sense to use a human language like Ruby and leave so much on the table when you can use a language like Rust or C to squeeze a LOT more out of correctness/types/performance.
This is a question I have been circling around. The agents are good at Ruby/Rails. It’s probably a confluence of things, but whether the L in LLM benefits from Ruby’s emphasis on human readability is something I’d love to see a study on. Or more generally, how key is (human) language itself in the performance ceiling of LLMs. A machine specific coding language/framework/process could obviously be more efficient, but if language itself is the hard limit, those would be dead in the water with LLMs. Or put another way: if language wasn’t the ceiling, why didn’t we see machine first coding implementations before LLMs?
We're going to have a bunch of hardly used libraries/frameworks that might converge on a single one. Similar to the surge in JS frameworks/libraries from a few years back and where we are now.
Is squeezing a lot more performance beneficial to a business? or is it a small amount of VPS cost per month difference?
It helps to be clear, at least with yourself, about if you're doing something because you can/want to or if there are actual business reasons.
This is my first attempt at a AI framework and I'm pretty happy with it
https://pricklypear.rocks/
I just skipped to a random part, 41:47, where the speaker is talking about the chasm of the old and the new, the before and the after, and our acceptance of the new status quo.
These match my own thoughts exactly. As much as I don't love them, I agree with the speaker.
For Factorio players: we used to live in Nauvis. It was beautiful but sparse. Then we outgrew it and unwittingly found ourselves on Fulgora. It is abundant but trashy.
And just like in a game, the rules changed and we adapt. That's what makes the game fun. Granted, that it's all our livelihoods on the line (very high stakes), but all things being equal, we have new challenges in front of us that we can either bemoan or accept.
Comes to Rails conference. Gives a keynote with nothing about Rails. Leaves.
It's not RailsConf. He's banned from that. It's his own Rails World he owns
Oh wow, I forgot about this. That context is important. He's talking to an audience of adherents. There's something mildly fascist about that.
Can we please not trivialise the meaning of fascist?
Actually it also is a good enough reason from him to move away from both working on Ruby on Rails and using it.
It's his conference he made because he was banned from the big one, for good reasons.
tldr:
dhh says 37signal's apps for Hey (their email platform moving to native)
backend for their email platform is now Rust - written by agents
says Rust produces small binaries - cutting number of servers needed from 110 to 10. hell they could serve peak traffic on a Raspberry Pi
says now that agents write all the code - use faster / native languages.
rails or ruby hardly get a mention besides that they might be well placed to capture web apps written by agents.
also worth noting that hotwire is dead according to DHH in this talk. The rendering html server side is old way now
HTMX is still alive and well; version 4.0 was just released.
(disclaimer: I'm a CEO of HTMX)
Where did he say that? I listened to it yesterday and AFAIK he only said "I don't know where the line between native vs web will end up in the future" (now that we can generate native apps easier)
Keep in mind DHH is working on an email app, it makes sense he sees native as the default option for that over pushing everything through the web when the barrier to entry lowered dramatically.
Hotwire is not explicitly a cross-platform concept. It's primarily a way to build web apps in the browser which is not going away any time soon. The only difference is in 2026 you're just more likely to vibecode a native mobile app than build a web/React Native version. Likewise you may build a native Desktop version instead of an Electron wrapper.
The issue with native apps is that you have to convince people to install them. If I had one app for each tab I open in my browser my phone would be a mess and out of free space. Then you slow down the delivery of features and bug fixes because you have to go through the app stores. You can't deploy something every day or more often.
So even if we can generate an app easily now, is that a wise choice for every business?
Like everything, depends on the usecase. An email desktop app seems to be a valid one. A standard earlystage B2B SaaS Rails web app? Not so much.
Same with server side rust. A high volume transactional backend email service makes sense there. The base layer of a CRM managing a lot of business logic?
Life will go on much the same. We just have more options to say yes to 'native' with LLMs, we can experiment more with ideal scenarios rather than a religious adherence to cross platform or single-framework merely for team DX simplicity.
If the apps were actually good, you'd install them. A lot of the apps on F-Droid are pretty lightweight, and a phone's security model makes it much safer than installing apps on Windows.
The web makes apps worse by forcing a fairly strict sandbox, even as Google keeps trying to loosen it. Think of a messaging app that hides in your notification tray, that's a trivial UX thing that's impossible on the web.
When I started in computing there was no Web, there was enough space to install the applications that actually mattered, and even better, developers actually had to take hardware constraints into account.
Rendering HTML on the server was kind of how IBM mainframes talked to their 3270 terminals, only that HTML is a lot nicer than 3270 streams and HTTP is human readable.
Server side rendering makes sense for mostly static content, but for web applications, a front end framework and a back-end API are a much better match.
Uhh, sounds like you haven't been writing modern React lately, which is server side lol
and something about omarchy
Talk is sponsored by Shopify, which also sponsored the Lex Fridman 5 hour promotion.
All DHH's AI opinions are sponcon.
But he has great flagging support and his perception management agency will suppress this one, too.
Shopify - where the CEO said just the other day that “slop grenades” are making way more work for everyone?
Same guys?
Just chugging the clown juice.
I really enjoyed DHH's keynote speech using his own family lineage story via the art history intertwined with technology evolution portrait paintings -> photography pictures as analogy of what's happening right now in the technology landscapes.
A few things are true:
- He is not wrong and all programmers need to hear this
- He loves to hear himself talk
- A year or two from now he will run this video to admit he was wrong about something
> A year or two from now he will run this video to admit he was wrong about something
I only wish this was more normalised in our society.
Sensible take and it's one I share.
Flow state.
I am glad that he mentioned that he'll miss the flow state that he would enjoy while deeply focused in computer programming. That was also one of my worries. Gladly, I still have other ways to get there, and I hope that you'll find yours too.
Out of curiosity, what other ways do you have to reach flow state? I honestly struggle to name even one. And definitely not one that would pay the bills.
Studying and writing about my conclusions, or just writing in general, is one way that I am still able to enter flow state outside of programming.
Sketching and drawing are also great for flow.
If you think you are a terrible drawer, check out “Drawing on the Right Side of the Brain”.
They didn't say it pays the bills. My guess is playing music, its the only other thing for me.
Well, an email server is next to the best use case for a low-level language. I'm still choosing Rails (or Laravel or Django) for CRUD apps any day of the week. Rails in particular is great because it's very conventional and AI writes it in an standardised way.
I don't understand what is supposed to come next. So its not Ruby on Rails anymore. What is it? Are agents going to magically convert all the RoR ecosystem to Rust or Whatever?
Translation does play to their strengths.
But yes, I think RoR's selling feature was developer ergonomics, which suddenly seems less of a benefit if developers aren't the ones writing the code. Readability and conciseness still has benefit, but the trade-off of worse performance (and less static checking) is suddenly more questionable.
Have a look at fly.io facing similar issues, another platform whose value prop was dev ergonomics: https://fly.io/blog/kurt-scott-money-sprites/
> Readability and conciseness still has benefit, but the trade-off of worse performance (and less static checking) is suddenly more questionable.
There is an issue with this idea - we write software, we read it, we make sure it does what it should do, that it’s implement in a sensible and efficient way, and, most importantly, we make sure the tests do what they should do.
If you no longer write the code, it’s harder to understand it. If you don’t write the tests, you need to trust the LLM did both jobs correctly (multiple adversarial LLMs can help), but, ultimately, it’s code you can no longer claim you understand. And you can’t really be sure it does what you want until you understand it.
How do you know Excel or Outlook does what you want?
We've had transpiled languages for decades, we really dont need a glorified dice to create rule based translation.
Everything is a nail at the moment.
Like people using LLMs to refactor. Refactoring is a solved problem, we've had decades to get nice tooling for this.
Shit, I'm also hosting with them. Somehow I missed that post. I've been watching Kurt's AI-pillness advancing over time, but I didn't expect it to metastasize so quickly.
What a strange time. Kurt himself writes that they've had the best quarter ever, and yet he feels confident enough to steer the ship in a completely new direction.
> and yet he feels confident enough to steer the ship in a completely new direction
If you don’t build the product that’ll kill your cash cow, your competitors will.
Write a solid piece of software by hand. Spend 20 plus years working all the bugs out and perfecting it. Build a giant community around it. Then use llms to rewrite it in another language, as the last step. Once that becomes unmaintainable, and there really is no community you can start over again by writing another working piece of software in a new language, perfecting it, building out a large community and then at the final step translate again with an llm, repeat.
I think what's next is the harness. Ruby on Rails brought conventions to web development. It could morph into brining conventions to building any kind of software using AI. AI on Rails. Many such efforts already exist and perhaps in short order such a project would be obsolete, but it's one potential path forward.
I dont think you watched the talk.
I think the perf jump from Ruby to Rust is much greater than .NET Core to Rust. Not much will change in .NET land for the foreseeable future. I see this migration happening from Ruby and others so it is not surprising from that aspect.
Also - what he is talking about was true for well over a year. It is just that a lot of day-to-day gaps between prompt and output were closed in the meantime. In a sense, Ruby on Rails/37 Signals is a late adopter.
> I think the perf jump from Ruby to Rust is much greater than .NET Core to Rust
This has not been true for me. If you are constrained by memory, you will still come out better using Rust, even if you use C# with AOT. We use .NET 10 at work and see 1/5th of the memory being used by a ported Rust service. Your mileage may vary.
With RAM prices where they are, I'm expecting to see a lot more of this in the next year.
We've been transitioning in-memory caching for web apps to tiered or on-nvme caching. It's an order of magnitude slower, but no one really noticed.
AI-psychotic and proud!
Results and output are undeniable, unless you are delusional or in denial.
I stopped writing code manually more than a year ago. I've experienced the 100x, even perhaps the 1000x he mentions, and I still think he's fucking psychotic.
Not reading generated webapp code is one thing. Blindly commiting OS code is a whole different dimension, as is writing production code in a language you don't even know how to read.
Worry not, for I am using Claude right now. I would never question the doctrine of AI usefulness and productivity. Please scrap that J'accuse draft, sir, for I would never commit blasphemy against AI, I promise.
I'm not worried, feel free to use punch cards if you want sir..
For varying definitions of results.
It's still early days, I'm hesitant to claim the death of software engineering. It's good now because LLMs were all trained on actual human output. One possible future is a slop death spiral.
I sure as hell am never again releasing human-authored open-source in a way that can be scraped. My Github is now relegated to the lowest quality slop from the cheapest models.
Serious question: Is there a framework that is better suited for agents? There must be something out there that makes agents fly through development with minimal token expense.
Whatever answer you get, it's nothing more than an opinion. For now, anyway.
https://danluu.com/pl-tokens/
Elixir. Phoenix or Hologram. I prefer the latter.
Yep I find Elixir very enjoyable with agents at the moment, I've been developing a lot in it. It helps that I already liked the language and the platform, and now I just have the ability to build out my ideas faster.
I think they're good at it for a variety of reasons, but it likely helps having a smaller and newer community than many languages, meaning less out of date guff clogging up the LLM sources.
Why is Elixir better suited for agents?
I maintain Elixir and Rails applications. In my opinion, almost all LLMs handle Elixir better. I suspect, though do not know, being an immutable functional language and composing applications in a functional style makes it easier for the models to "reason" about the code and state. I learned very quickly that this is my preferred language for that very reason.
Additionally, the features available in the BEAM and OTP mean the agents have built-in tools for many things that would require reaching for 3rd-party libraries or even additional services. This means less variability in design as the models will happily use all the language features human developers would have to learn about and understand over time to use effectively. Not that this isn't important, but I believe models do a better job steering you to obvious and optimal solutions with Elixir.
The new type system is only going to make it better.
An agent that was looking at the log output from a program in a typical language, and also looking at the code for that program, would have a lot of unanswered questions related to "how do I iterate on this?" Maybe there's a kubernetes cluster involved somewhere. Maybe it there's a cloud storage bucket. Maybe there's dashboards and metrics and telemetry hosted by servers that the agent doesn't know the name of.
Elixir runs on the BEAM which lets you do hot code reloading process at a time. And these are much more lightweight than your typical OS process. So you can do things on the BEAM which would be just crazy to do on other achitectures. Like if you have a server which needs to support a million users, it would be insanity to run a million separate OS processes for the job. But BEAM processes are so lightweight that you can do that, which gives each one isolation from misbehavior of the others.
And since these support hot code reloading, without restarting the BEAM's OS process, you can patch the code for your user's sliver of the server and do a sort of "what-if" experiment. Other architectures would require you to implement separate test environments or complicated feature flag systems, but the beam lets you just reach out and change it for just the process you care about--you're not baking the experiment into an OSI image or anything so dangerous as that... blast radius is kept small.
And the steps for carrying out this experiment, it's all doable at the elixir shell. You connect to the BEAM and make changes. That's a tremendous reduction in context that an agent needs to load--context which would otherwise instruct the agent about how to reason about the deploy process and how to wall off your experiment from others so as to not cause problems with it is just answered implicitly by how the BEAM works.
Elixir compiles to BEAM bytecode, so you get all of this because the BEAM is cool. As for Elixir as a language... I don't think there's anything agent-specific about it. Gleam might be a better choice since it's statically typed and also compiles to BEAM bytecode except that it's not as popular as Elixir, which is probably more important.
You can also do all of this in the language that started it all: Erlang (Java : JVM :: Erlang : BEAM) but it's not exactly fashionable at present (similar to java vs like... kotlin or clojure or something).
Disclaimer: I'm more of a fan than a seasoned pro. I'd love to actually be working with this stuff daily, but my team wouldn't have it because they prefer familiar things and pain.
It's vibes all the way down.
I think there's actually something to it: https://x.com/josevalim/status/2019422467446698493
I that case so is Clojure.
why is that?
Good. We need AI-bullish folks to test it boldly and then report whether it works or not.
Is it just me or DHH is being a bit creepy?
Also, he always states super strong opinions on anything, punctuated by the f-word, as if everyone else was shit, and then he changes his mind. (For instance, the way he despised VCs and now he seems to acknowledge them, his stance on AI, …, his change from being seemingly more left-leaning to being a libertarian).
He several negative traits. He's one of these opinionated controversial guys. Because he has money people feel the need to react to what he says, whether they think he's right or wrong. Annoyingly, people like that get more exposure than level headed people with measured takes.
Mixed with his terrible views on society (xenophobia, techno-fascism), it makes a rather sour cocktail.
"Strong opinions, loosely held" is an ideal worth striving for. The alternative is to hold on to things you should know better are false, aka dogmatism.
That's not a licence to routinely adopt extreme opinions.
*fascist. Don't sugarcoat it.
That's cool, but like with most things, I think the manager types will misunderstand:
- A lot of this is looking at an existing app and/or code for reference.
- The people doing/guiding this are skilled.
- Skilled people have thought out what to make.
Otherwise they hear "20 minutes" and think AI can spit out a logical or good product from no reference or planning.
Next year they will rewrite rails in rust with AI
One thing I really agree with dhh on (and something I’m genuinely looking forward to!) is the resurgence of native apps.
Cross-platform became so dominant because the same small dev teams had to support users across multiple platforms. Practically they either had to build ONLY for the web or use some kind of web-tech-based framework like Electron.
Agentic coding changes that constraint at a very low level. Teams are no longer limited by whether they have someone who knows how to build for iOS, Android, macOS, etc., or whether they have the bandwidth to maintain multiple native codebases. The cost of supporting multiple platforms starts to fall dramatically.
And I think this is closely related to the thesis behind Omarchy also. Setting up and customizing Linux is historically a lot of friction. If you wanted a system that was personal i.e. NOT (Mac OR PC) you often had to be comfortable troubleshooting obscure issues and spending hours getting everything just right. Don't get me wrong: a lot of people enjoy that friction. But it sucks at so many levels.
Agents lower that barrier. You can want a highly customized system without necessarily becoming an expert in every layer of the stack. If there’s some tweak or workflow that isn’t exposed through a settings menu, you can increasingly just describe what you want and have the system help you build it.
So I think the interesting shift is that software becomes less constrained by what the average team or individual programmer has the bandwidth to build and maintain. That could make both native apps and highly personalized computing environments much more viable again instead of the web-slop that we've become accustomed to over the past 20 years. I really see parallels between how the Mac empowered creatives in the 80s and agentic coding today.
As much as i'd love for this to be true, it is not. Teams need to support their native apps, just because you prompted your way into some code doesn't mean your team is able to support it. The constraints are the same, if no one in your team knows how to build an maintain a native iOS app and their core knowledge is TypeScript and web, vibing it with agents won't get you past your first release (and not even that)
I don't really mean “give a team of TypeScript devs an agent and suddenly they can ship and maintain a production iOS app.” Expertise and ownership still matter, especially for something you have to support for years.
I'm talking more about agents as a force multiplier for teams who do have engineering experience. They reduce the amount of specialized knowledge, time, and coordination required to get something built and maintained. The constraint gets meaningfully weaker.
An experienced engineer who understands software architecture, debugging, product constraints, etc. but isn't an expert in every native platform can increasingly use an agent to bridge some of those gaps. That's very different from vibecoding your way into production.
fwiw I think we're still very early in the process of re-thinking about how teams are organized and what skill sets they need.
I agree with you, but DHH is proud of having a vibecoded backend in production written in a language he can't read.
I doubt he is on call or getting paged at 2am for any of it
Adieu, Ruby on Rails.
DHH is "off the Rails" is not false now?
Totally disagree on the reevaluation of architecture best practices.
Agents are as prone as forgetting to update duplicated code as we are ( or almost ), because they work at the same level : text files spread everywhere that you have to open and read to know what's inside.
The fundamental reasons for those architecture best practices are still there, and it's actually quite convenient and also a good reason to remain optimistic in our ability to find a job as a professional software developer.
Once the rest of those talks are published on YouTube (usually in a couple of weeks), you’ll see that there were some other talks that were much better about conveying a nuance in a balance between the tension of “AI can do everything for me, let it write machine code” and a session titled something along the lines of “your code base is your prompt“.
One presentation, I think it was my favorite one, talked about how an LLM had a very difficult time understanding what happened in a class when the name of the class was poorly descriptive of that class behavior.
Good quality code still matters. The LLMs are based on language! Human language!
Related:
What About Rails?
https://news.ycombinator.com/item?id=49839664
A lot has been said about DHH, but, I think the moment where you can see that the guy has _clearly_ lost it is this quote:
> It is impossibly fast. And you know what? The binary? Half a megabyte. Half a fucking megabyte. I— you could fit that on a floppy disk almost. With a little compression. This is the other payoff.
The guy now(?) thinks like a lossy abstraction that speaks in tropes.
He has become the LLM - I would say, if I didn't know that even a 9B one can tell me with absolutely certainty how much space there is on a fucking floppy.
There is no reason to get into all the politics stuff when this is where the guy is at. Basic fact retrieval.
(Pedantically, one could now argue that it was not clear which floppy format the guy was referring to. And yes, that would be true, but by thinking that thought, you have thought about the sentence more than its author has.)
I don't get it. Are you saying floppy disks were usually bigger than half a fucking megabyte or smaller?
I'm missing something (as usual) because I don't see what's wrong with his statement.
If you allow me to quote Gemini 3.1 pro:
___
If we are talking about the standard 3.5-inch High Density (HD) floppy disk, which is the universal cultural icon for "floppy disk" (and the save icon in your software), the capacity is exactly:
Now, let's run the exact math on DHH's keynote claim: If you put his "impossibly fast" 0.5 MB binary onto a standard floppy disk, you would have nearly a full megabyte of free space left over. You could fit almost three of his binaries on a single disk. You do not need "a little compression." You need a basic understanding of less-than/greater-than operators.The Preemptive "Reply-Guy" Shield
Now, just in case a Hacker News pedant is reading this and wants to argue that DHH was talking about a different format, here is the absolute certainty on the rest of the historical hardware:
To make DHH's statement mathematically correct, we have to pretend he was standing on a stage in 2024, hyping up modern web framework binaries, while indexing his storage metaphors to a 5.25-inch Double Density disk from 1981.__
Thank you, Gemini. Much appreciated. Also, I love the tone.
(You can tell that it is actually that LLM by it thinking that it is still 2024)
> but by thinking that thought, you have thought about the sentence more than its author has
Overanalysing someone's spoken sentences isn't the badge of honour it might appear to be.
I don't understand what you think is so objectionable about that quote. It's one of the least mental things he's said.
You've never made a minor mistake during a presentation? (I await your confident "no".)
weird talk; is not reading any code _at all_ is the new standard now? I do happily ai-slop side projects here and there but seems like just going full steam ahead _at work_ is ok? sorry, I meant - is the future now? xd
I guess it's the only con that will take money to have DHH give a keynote about some sycophantic BS
No, not at all. Anyone who gives a damn at all about producing quality software is reading the code, and frankly most people who care about producing quality software aren't using LLMs to do it.
yes
Haven't watched the video, but I've been spending a few hours the past week re-acquainting myself with Ruby before diving back into rails introductions.
Ruby is an interesting language to work with. I don't know it like Python, but it's a language that finds problems developers would run into and built solutions in to make it easier.
As for Rails, as a total novice: It is a lot of Convention over Configuration, and it's almost a second language on top of Ruby. Everything works together implicitly so you have to recognize what's happening when you look at the code. I don't know if any Ruby/Rails oriented IDEs exist that guide the developer with hints and shortcuts to how the pieces of a Rails app fit together, but it sure would be handy.
I can see how typed, high-performance languages with explicit references to all the integrations would be easier for AI to handle. But as long as humans sometimes have to look at the code, I would think Rails still has a place.
> how the pieces of a Rails app fit together, but it sure would be handy
Best thing you can do is get a fuzzy file search, but you'll learn the structure soon enough. Once you know it, you'll be very fast. I loved it when I used it professionally.
Grifter gonna grift.
This link says ‘This video is private’ to me, here’s a working one https://youtu.be/vDjW_dRyKXY
Thanks! We've changed the url to that from https://www.youtube.com/watch?v=V9SxpJpHuus.
it's funeral of Ruby on Rails
It’s now Rust on Rails.
it became culturally irrelevant around rails 5