I think what’s interesting about AI, and why there’s so much conversation, is that in order to be a good user of AI, you have to really understand software development. All the people I work with who are getting the most value out of using AI to deliver software are people who are already very high-skilled engineers, and the more years of real experience they have, the better.
I know some guys who were road warriors for many years —- everything from racking and cabling servers, setting up infrastructure, and getting huge cloud deployments going all the way to embedded software, video game backends, etc. These guys were already really good at automation, seeing the whole life cycle of software, and understanding all the pressure points. For them, AI is the ultimate power tool. They’re just flying with it right now. (All of them also are aware that the AI vampire is very real.)
There’s still a lot to learn, and the tools are still very, very early on, but the value is clear.
I think for quite a few people, engaging with AI is maybe the first time ever in their entire career they are having to engage with systems thinking in a very concrete and directed way. Consequently, this is why so many software engineers are having an identity crisis: they’ve spent most of their career focusing on one very small section of the overall SDLC, meanwhile believing that was mostly all there was that they needed to know.
So I think we’re going to keep talking for quite a while, and the conversation will continue to be very unevenly distributed. Paradoxically, I’m not bored of it, because I’m learning so much listening to intelligent people share their learnings.
> I think what’s interesting about AI, and why there’s so much conversation, is that in order to be a good user of AI, you have to really understand software development.
This I agree with completely. You can see it in the difference between a prompt where you know exactly what you want and when things are a little woolley. A tool in the hands of a well trained craftsperson is always better used.
> So I think we’re going to keep talking for quite a while Me neither, and to be clear I'm okay with that. This was mostly a rant at the lack of diversity of discourse.
Agree, the diversity of the discourse is not great. There's a lot of "omg I just got started waaauw" articles out there along with "we're all gonna die!" stuff. And then a few seams of very excellent insight.
Deep research at least helps with dowsing for the knowledge...
{heart}
https://www.youtube.com/watch?v=Q0iuMByFjCQ&list=PLEouLkiLHd...
This isn’t to say there’s not hype. Just that if you’re not seeing big productivity gains you need to make sure you really are an outlier and not just surplus to requirements.
Rather, I hear a lot of nuanced opinions of how the tech is useful in some scenarios, but that the net benefit is not clear. I.e. the tech has many drawbacks that make it require a lot of effort to extract actual value from. This is an opinion I personally share.
In most cases, those "big productivity gains" are vastly blown out of proportion. In the context of software development specifically, sure, you can now generate thousands of lines of code in an instant, but writing code was never the bottleneck. It was always the effort to carefully design and implement correct solutions to real-world problems. These new tools can approximate this to an extent, when given relevant context and expert guidance, but the output is always unreliable, and very difficult to verify.
So anyone who claims "big productivity gains" is likely not bothering to verify the output, which in most cases will eventually come back to haunt them and/or anyone who depends on their work. And this should concern everyone.
The only downside is not learning about method Y or Z that work differently than X but would also be sufficient, and you don’t learn the nuances and details of the problem space for X, Y, and Z.
I've learned about outbox pattern, eventual consistency, CAP theorem, etc. It's been fun. But if I didn't ask the LLM to help me understand it would have just went with option A without me understanding why.
I am amazed at the incredible things it can do - only to turnaround and not be able to do a simple task a child can do. Just like people.
But at the same time the more I read about AI, the more I realize I need to learn about AI. Thus far I'm just using cursor and the Claude code extension alongside obra superpowers, and I've been quite happy with it. But on Twitter I see people with multiple instances of Claude code or open claw talking to each other and I don't even know how to begin to understand what's going on there. But I'm not letting myself get distracted — Claude code and open claw are tools. They could go away at any time. But systems thinking is something that won't go away. At least, that's my gambit.
I don't mean to snipe at AI, because it really does seem to have set more people on the path of learning, but I was writing VB5 apps when I was 14 by copying poorly understood bits and pieces from books. Now people are doing basically the same but with less typing and everyone thinks it's a revolution.
I think it is genuinely impressive to be able to build one app with AI. But I haven't seen evidence that someone could build a maintanable, scalable app with ai. In fact, anecdotally, a friend of mine who runs an agency had a client who vibe coded their app, figured out that they couldn't maintain it, and had him rewrite the app in a way that could be maintained long term.
Again, I'm not an Ai detractor. I use it every day at work. But I've needed to harden my app and rules and such, such that the AI cannot make mistakes when I or another engineer is vibing a new feature. It's a work in progress.
I don't believe they said that folks new to AI can't make impressive use of it. They did however say that senior folks with lots of scrappy and holistic knowledge can do amazing things with it. Both can be true.
They still have absolutely no clue how it works, so how could they "write entire applications"? They vibed it, but they certainly didn't write any of it, not one bit of it, and they're clueless as to how to extend it, upgrade it, and maintain it so that the AI doesn't make it a bloated monstrosity of AI patches and fixes and workarounds that they simply could never begin to understand.
They were also following a dozen youtube tutorials step by step, so even that part was someone else doing the thinking.
Yeah, these are the same guys constantly bugging me to help them figure something out.
AI makes a ton of bad decisions too and it's up to you to work with it. If I had the knowledge of the dangers hidden in things I'm developing, I'd move even faster
Was able to make a great full web app, which I think is hardened for prod but it had to be refactored to do so. Which it happily did.
It's really about asking the right questions, breaking down tasks, and planning now. I'm going to tackle a huge project, hoping to share it here.
Well, there was also a lot of unrelated things that happened as well around last November for me, but yes, getting into vibecoding for real was one of them, and man I feel physically drained coming back from work and going to use more AI.
Not sure what it is. I'm using AI personally to learn and bootstrap a lot of domain knowledge I never would have learned otherwise (even got into philosophy!, but man is it exhausting keeping up with AI. I would burn through a week's worth of credits in a day, and now I haven't vibe coded a week.
I think, I will chill. One day at a time.
My take is that it's similar to what Amber Case described in Calm Technology - with AI you are not steering one car, you're really steering three cars at the same time. The human mind isn't really designed for that.
I am finding that really structuring my time helps in terms of fighting back. And adopting an hours restriction, even if I could rage for 4 more hours, I don't. Instead I stop and go outside.
Me too. A key purpose of HN, and a bright time for that.
Now I'm working on a second project, all with AI. I haven't written a single line. It works better than a non programmer would make because I knew what to ask for. But I'll admit I'm not learning anything.
Then the same happened with languages that managed memory.
And with IDE that could refactor your code in a click and autocomplete API calls.
And with Stack Overflow where people copy/pasted code they didn't understand.
So yeah i mean - who cares how it works - but also if you have experience in how things _do_ work you can solve problems other people cannot.
For example, I haven’t racked and cabled a server in over 15 years. That used to be a valuable skill.
I also used to know how to operate Cisco switches and routers (on the original IOS!). I haven't thought about CIDR and the difference between a /24 and a /30 since the year 2008. A class IP addresses, how do those work? What subnet am I on? Is thing running on a different VLAN? Irrelevant to me these days. Some people still know it! But not as many as in the past.
The late Dr. Richard Hamming observed that once a upon a time, "a good man knew how to implement square root in machine code." If you didn't know how to do that, you weren't legit. These days nobody would make such a claim.
So some skills fade and others rise. And also, software has moved in predictable cycles for many decades at this point. We are still a very young field but we do have some history at this point.
So things will remain the same the more they change on that front.
Also anyone making a homelab has to know these stuff.
And there'll be a split too... like there's a giant divide between those mechanics who used to work on carburetors and the new gen with microcontrollers, injection systems, etc. People who think cars are 'too complicated' aren't wrong, but for someone who grew up in the injected era, i vastly prefer debugging issues over the canbus rather than snaking my ass around a hot exhaust to check something.
I don’t know if it’s the Universe delivering this farce or it’s the emergent LLM Singularity.
That's not what I said. I said that those who are already shining, are now shining even brighter. Give a great craftsman a new tool and he will find a way to apply it. If it is valueless, he will throw it away.
For what it's worth, your comment is also an HN trope, the disaffected low-effort armchair keyboard warrior.
It's what happened with the internet and computer usage. As Apple made it easier to get online with zero computer knowledge, suddenly we're electing people like donald trump.
AI is the thing that for the first time can think better than us (or so at least some people believe) and is seen as an efficiency booster in the world of cognition and ideas. I'd think Hannah Arendt would be worried with what we are currently seeing and where we might be headed.
Turns out Lowtax was right and ahead of his time
We have hundreds and thousands of years of history showing humans committing atrocities against each other well before the advent of computers, or even the introduction of electricity. So while the tool may become so ubiquitous that there’s no option not to engage with it, I don’t think it really fundamentally alters the dynamics of human behavior.
Some people are motivated by greed. Others are motivated by nobility. It really just comes down to which wolf they're feeding.
In terms of the tool keeping people ignorant, there’s a part I agree with and a part that I don’t. I think, in terms of information dissemination, AI is probably the autocrat’s wet dream in terms of finally being able to achieve real-time redefinition of reality. That’s pretty scary, and I’m not sure what to do about it.
On the other hand, people have always been free to not really learn their craft and to just sort of get by and make a living. That was true a thousand years ago, and it’s true today. There’s always somebody who can do really a high-quality job, but they’re very expensive, and then there's a vast population who will do a medium to terrible job for less money. You get what you pay for. There's a reason history is primarily written about people with power and wealth, they were the only ones with the means to do anything.
I don’t agree with the assertion about the internet and the election of someone like Donald Trump. Well before the internet existed, politicians were using communication mediums to influence things and get elected—whether it was the telegraph, the telephone, or the TV. JFK famously was the first TV president (notably, he didn't wear a hat).
These technologies simply give politicians more reach, and they may change the dynamics of how voters are persuaded. But what’s true today was true three hundred years ago: there’s the face of power that you see publicly, and then there’s what really happens behind the scenes.
Spoken like someone who thinks they are going to be insulated from the fallout
Sure, it might hurt me personally. I'm not selfish enough to put that over what will be an incredibly empowering development for our species.
This will be good for a handful of elites and no one else
Where are they flying and why software has gone to shit?
Maybe this super stars programmers have to keep their reality breaking technology secret, but everything has not only degraded, but turned to absolute trash.
Yup
AI seems great when you have no way of truly validating its output.
My partner teaches at a small college. These people are absolutely lost, with administration totally sold on the idea that "AI is the future" while lacking any kind of coherent theory about how to apply it to pedagogy.
Administrators are typically uncritically buying into the hype, professors are a mix of compliant and (understandably) completely belligerent to the idea.
Students are being told conflicting information -- in one class that "ChatGPT is cheating" and in the very next class that using AI is mandatory for a good grade.
Its an absolute disaster.
It is all small things, but none of those small things are captured anywhere so whoever is on the other end has to 'discover' through trial and error.
The closer they can map their real problems to make-document-bigger, the better their results will be.
Alas, that alignment is nearly 100% when it comes to academic cheating.
This is the opposite.
They paid for custom on prem software and in over a year, they have not fully provided both access and infrastructure for install it.
We have been paid already, but they paid for a tool they can’t get their shit together enough to let us install.
Doesn't sound that different from my tech job
So in most courses nothing has changed in the way we grade. Suddenly coursework grades have gone up sharply. Anyone with working neurons know why, but in the best case, nothing of consequence is done. In the worst case (fortunately uncommon), there are people trusting snake oil detectors and probably unfairly failing some students. Oh, and I forgot: there are also some people who are increasing the difficulty of the coursework in line with LLMs. Which I guess more or less makes sense... Except that if a student wants to learn without using them, then they suddenly will find assignments to be out of their league.
So yeah, it's a mess.
My son, who is a freshman at a major university in NYC, when he said to his freshman English professor that he wanted to write his papers without using AI, was told that this was "too advanced for a freshman English class" and that using AI was a requirement.
Now I have this love/hate relationship with it. Claude Code is amazing. I use it everyday because it makes me so much more efficient at my job. But I also know that by using it I’m contributing to making my job redundant one day.
At the same time I see how much resources we are wasting on AI. And to what end? Does anybody really buy the BS that this will all make the world a better place one day? So many people we could shelter and feed, but instead we are spending it on trying to make your computer check and answer your emails for you. At what point do we just look up and ask… what is the damn purpose of all of this? I guess money.
I know someone who worked for a nonprofit that made pregnancy health software that worked over text messaging. Its clients were women in Africa who didn’t have much, but they had a cell phone, so they could get reminders, track vitals, and so forth.
They had to find enough funding to pay several software engineers to build and maintain that system. If AI allows a single person to do it, at much lower cost, is that bad?
So in isolation, I think it's great that they managed to achieve this. But I mourn that the only way they achieved it was via this rapacious truth-destroying machine.
This isn't a new trend - AI didn't cause it. It's just the latest version of it.
The actual community building is fairly not as automated unless you have very specific problems. Like even in the example above, having an automated message is useful but staffing the team to handle when things are NOT in a good spot would probably be the real scaling cost.
To add to list of questions - it's undeniable the AI is making humans dumber by doing mental job previously done by humans. So why we spend so much energy making AI smarter and fellow humans dumber?
Shouldn't we be moving in opposite direction - invest in people instead of some software and greedy psychopaths at helm of large companies behind it?
I don't see how this is the case if you're anything more than a junior engineer... it unlocks so many possibilities. You can do so much more now. We are more limited by our ideas at this point than anything else.
Why is the reaction of so many people, once their menial work gets automated, "oh no, my menial work is automated." Why is it not "sweet, now I can do bigger/better/more ambitious things?"
(You can go on about corporate culture as the cause, but I've worked at regular corporations and most of FAANG. Initiative is rewarded almost everywhere.)
> Does anybody really buy the BS that this will all make the world a better place one day?
Why is it BS? I'm shocked that anyone with a love and passion for technology can feel this way. Have you not seen the long history of automation and what it has brought humanity?
There is a reason that we aren't dying of dysentery at the ripe age of 45 on some peasant field after a hard winter day's worth of hard labor. The march of automation and technology has already "made the world a better place."
because i have rent to pay? old age to prepare for?
why is it so hard to understand most people are not rich, that the cost of living is high, and that most people are VERY afraid their jobs will be automated away? why is so hard to understand that most people haven't worked at FAANG, they don't have stocks or savings, and are squeezed harder with every new day and every new war?
what world, what reality are you guys living in?!
"Software engineer" as a profession is rapidly getting automated at my company, and yet our SWEs are delivering more value than ever before. The layer of abstraction has changed, that is all.
> what world, what reality are you guys living in?!
One that has seen immense benefits from the Industrial Revolution and previous waves of automation.
Do you think because 2 dev are now super productive with AI, the company will keep the other average 30 devs? no, of course not, they will fire and pocket the difference. Same for other industries, where AI will slowly diffuse like a poisonous gas and displace jobs and people, leaving behind a crippled white collar class. The profits will not trickle down and the increased productivity will be a hatchet, not a plough.
Yes, they will keep the other devs that can figure out how to use AI well. Businesses want to grow.
Such things were super uncommon before the industrial revolution, I'm sure.
I really don't understand this way of thinking. Don't you think that AI could replace senior engineers? Sure, companies will be able to do bigger / better / more ambitious stuff - but without any software engineers.
> Why is it BS? I'm shocked that anyone with a love and passion for technology can feel this way. Have you not seen the long history of automation and what it has brought humanity?
I definitely think that AI will be a net benefit for society but it could easily end up being be bad for me.
I use AI agents every day at work and I'm happy with that, but it took over two years and billions of dollars in investment to deliver anything useful (Claude Code et al). The current models are amazing, but they still randomly make mistakes that even a junior wouldn't make.
There's another paradigm shift to be made certainly, because currently it feels like we scaled up a bug brain to spit out code. It works great for some problems, but it's not what software developers usually do at work.
the swe role is going to change but problem solving systems thinkers with initiative won't go away
We owe it to the world, as the experts, to be critical. The march of automation and technology has made the world a better place in some ways. I sure love modern medicine, but those drones flying over Ukraine and Russia sure don’t seem like they are making the world a better place. Nuclear bombs are not making the work a better place. Misinformation in social media is not making the world a better place.
Any belief you drink blindly will eventually find a way to harm you.
Strangely enough, I don't see you calling to end the consumption of meat which would have a far larger environmental impact while not slowing global progress at all.
Tech is what got us where we are. AI allows us to use more energy to produce more of what is currently measurably killing us.
> but AGI might save us from it.
This is just faith. Some believe that prayers may save us.
Many things are orders of magnitude bigger than AI in the energy usage problem that bring less comparable value.
Tell that to the people who will die before 45 because of global instability and global warming, I guess?
Everyone is in their own place adapting (or not) to AI. The disconnect b/w even folks on the same team is just crazy. At least it's gotten more concrete (here's what works for me, what do you do) vs catastrophizing jobpocolypse or "teh singularity", at least on day to day conversations.
For a while, it felt like I'm in a minority when I was saying that it can be a useful tool for certain things but it's not the magic that the sales guys are saying it is. Instead, all the hype and the "get rid of your programmers" messaging made it into this provocative issue.
HN was not immune to this phenomenon with certain HN accounts playing an active part in this. LLMs are/were supposed to be an iteration of machine learning/AI tools in general, instead they became a religion.
> here's what works for me, what do you do
This is at least progress... but many want to remain in denial, and cant even contemplate this portion of the conversation.
We're also ignoring the light AI shines on our industry, and how (badly) we have been practicing our craft. As an example there is a lot of gnashing of teeth right now about the VOLUME of code generated and how to deal with it... how were you dealing with code reviews? How were you reviewing the dependencies in your package manager? (Another supply chain attack today so someone is looking but maybe not you). Do you look at your DB or OS? Does the 2 decades of leet code, brain teaser fang style interview qualify candidates who are skilled at reading code? What is good code? Because after close to 30 years working in the industry, let me tell you the sins of the LLM have nothing on what I have seen people do...
All that said, it’s extremely exciting. I’ve been in tech, in one way or another, for 25 years. This is the most energizing (and simultaneously exhausting) atmosphere I’ve ever felt. The 2006-2011 years of early Facebook, Uber, etc. were exciting but nothing like this. The future is developing faster than we can process it.
I write Typescript and SQL by day, my last two personal projects were Rust and Perl.
I do worry that I'm not learning them as deeply, but I am learning them and without AI as an accelerant I probably wouldn't be trying them at all.
We're about due for some new computing abstractions to shake things up I think. Those won't be conceived by LLMs, though they may aid in implementing them.
The stacks of turtles that we use to run everything are starting to show their bloat.
The other day someone was lamenting dealign with an onslaught of bot traffic, and having to deal with blocking it. Maybe we need to get back to good old fashioned engineering and optimization. There was a thread on here the other day about PC gamer recommending RSS readers and having a 36gb webpage ( https://news.ycombinator.com/item?id=47480507 )
(though it sounds like if you left it for long enough, you'd get 36 GB of ads downloaded eventually)
Saame. I wonder if the use of AI will lead to less invention and adoption of new ideas in favour of ideas with lots of training data.
But I don't see it that way. I've been fascinated by AI since I was a little kid (watching Max Headroom, Knight Rider, Whiz Kids, Wargames, Tron, Short Circuit, etc in the 80's) up through college in the 1990's when I first read about the 1956 Dartmouth AI workshop that kicked the field off, and up to today where we have the most powerful AI systems we've had. Every single bit of this stuff is wildly fascinating to me, but that's at least in part because I recognize (or "believe" if you will) that there's a lot more to "AI" than just "LLM's" or "Generative AI".
I still believe there are plenty of neural network architectures that haven't been explored yet, plenty more meat on the bone of metaheuristics, all sorts of angles on neuro-symbolic AI to work on, etc. And even "Agents" are pretty exciting when you go back and read the 90's era literature on Agents and realize that the things passing for "Agents" right now are a pretty thin reflection of what Agents can be. Really understanding MAS's involves economics, game theory, computer science, maybe even a hint of sociology.
As such, I still find AI fascinating and love talking about it... at least in the right context and with the right people. :-)
And besides... as they[1] say: "Swarm mode is sick fun".
[1]: https://static0.srcdn.com/wordpress/wp-content/uploads/2022/...
For me, the issue isn't that I can't conceive of work AI could help with. It's that most of the work I currently need to be doing involves things AI is useless for.
I look forward to using it when I have an appropriate task. However, I don't actually have a lot of those, especially in my personal life. I suspect this is a fairly common experience.
I'd imagine similarly there were points in time where people who go to concerts just to see the electric guitars and lighting setups.
Argueably Jean-Michel Jarre concets were 100% gear-porn shows.
news.ycombinator.com##td.title:has-text(/LLM|AI/i)[0] https://en.wikipedia.org/wiki/Technology_adoption_life_cycle
Alternately: the trough of disillusionment.
> What makes this worse, is our bosses have bought into it this time too. My managers never cared much about database technologies, IDE’s or javascript frameworks; they just wanted the feature so they could sell it. Management seems to have stepped firmly and somewhat haphazardly into the implementation detail now. I reckon most of us have got some sort of company initiative to ‘use more AI’ in our objectives this year.
I am extremely skeptical of AI products anyone builds. It's just using one black box to build scaffolding around another black box and then typically want to charge money for it. I don't see any value there.
Also, depends on who target user is.
AI can be used to build deterministic software
Before that we were excited about the wheel and the creation of fire. All capital drained into those ephemeral fancies.
The cycles cycle on.
Like the new frontend frameworks coming every week after 2010 sometime. Not jumping on every single one, and waiting until react was declared the winner and learn that worked well. Sure, someone that used it from day 1 had more experience, but one quickly catch up.
Only thing that stuck thus far is the cloud. Though not for infinite scalability and resiliency, cause that just dumps big invoices in your lap.
The Cloud happened as well, as you've pointed out
AI adoption is well past Quantum and Web 3. Comparing it to those two is nonsensical.
All those listed and more, are part of the cycles that the parent comment mentioned and which I've continued.
Same thing with Agile. Mostly sprint-based waterfall, iterative development is not something I've ever seen in practice. Or people over processes, remember those ideas?
BigData, was another hype cycle where even smaller companies wanted a "piece of the action". I've worked at the time in a sub 50 developers company, and the higher ups where all about big data. When in fact our system was struggling with GBs of data due to frugality in hardware.
For a moment in time you couldn't spit in any direction without hiting a Domain Driven Design talk. And now we disable safeguards and LLMs write a mix of garbled ideas from across all the laundered open source training data.
Too early to tell where AI will land, and if it will bring down the economy with it, but spending rate doesn't deliver equal results for all, and we will have to see after the dust settles.
---
Personally, I'm still very interested in the topic.
But since the tech is moving very fast, the discussion is just very very unevenly distributed: There's lots of interesting things to say. But a lot of takes that were relevant 6 months ago are still being digested by most.
Never heard this and I like it very much. This is just an off-topic comment to say thanks!
I don't like the hype language applied by the channel host one bit - and so this is not something where I expect someone tired of the hype to be swayed - but I think his perspective is sometimes interesting (if you filter through the BS): He seems to get that the real challenge is not LLM quality but organisational integration: Tooling, harnesses, data access, etc, etc. And so in this department there's sometimes good input.
The new HN is full of people filled with anxiety about being replaced by an advanced calculator.
To an outsider, it could almost be funny if it wasn't so sad.
[1]: https://en.wikipedia.org/wiki/Mastodon_(social_network)
Our local tech meetup is implementing an "LLM swear jar" where the first person to mention an LLM in a conversation has to put a dollar in the jar. At least it makes the inevitable gravitational pull of the subject somewhat more interesting.
All that said, I've already set up a few of my non tech close friends with Cowork and they are huge fans of it now. It's somewhat shocking how much menial repetitive work the average white collar job entails.
At my big tech, AI is every conversation with everyone, every day. Becoming AI native is a huge deal for us. Literally everyone is making AI usage a core part of their job and it's been a big productivity accelerator.
Perhaps it's different where you work, so you don't see the sentiment.
Wow that sounds horrible.
I spent 2024 on Mastodon and I absorbed their groupthink that AI was useless... I wish I could get that year of my life back. I wish I had that extra year headstart on AI compared to where I am now. So much of my coding frustrations that year that might have been solved from using AI. I am reluctantly back on X - I hate what has been done to Twitter, but that's where so much of the useful information on using AI is being shared.
Well, back to it. Claude has been building another local MCP server tool for me in the background.
100% feeling this divide as well.
People that deny the benefit of AI in 2026... I can't even engage with them anymore. I just move on with my life. These people are simply not living in reality, it will catch up to them eventually (unfortunately.)
I'm also learning art and I'll never use AI here, so I thought I have less time for hobby programming and I could just use AI for that but then I come back to the concern I mentioned above.Plus I can't proudly share anything I made with it either because I wouldn't have done much of the work at all.
I'm also feeling burnt out about web dev in general and doing the same thing during my free time just feels like more work to my brain. I wish I could find something interesting to do, and if I don't I'll quit programming in free time for good.
The new GenAI architectures and tooling supported by them just give more fun things to do and fun ways to do it.
All they essentially did was tell the LLM to test and verify whether the answer is correct with a prompt like the following:
>"You just edited X. Before moving on, verify the change is correct: write a short inline python -c or a /tmp test script that exercises the changed code path, run it with bash, and confirm the output is as expected."
Now whether this is true, I don't know, but I think talking about this kind of stuff is cool!
Endlessly grooming the Agent reminds me of Gastown.
Curios to see what he'll present, if, from his 700+ contributions in private repositories.
As shown in "Normal Accidents" the strength is as high as its weaknesses, and in any complex system this is even more a problem. A catastrophic event is still to happen with AI as it happened in basically every complex system. They ocurred with trained people that wasnt believing in magic or laziness... so the scenario is even worse for AI.
Yes, I'm bored about people that believe in magic and the ghosts the are emerging and are yet to be seen.
But nobody wants to hear about prompt calibration or pipeline architecture. They want to hear "I replaced my whole team with agents." The boring, useful work is invisible, and the flashy stuff gets all the oxygen
It will calm down once the dust starts to settle and there's some kind of consensus on how the chips have fallen.
Also there is an irony that talking about being sick of talking about AI is still talking about AI.
I think the "can do" part gets boring but now I'm paralleling this to trust relationships and fiduciary responsiblities. What I mean is that we can not only instruct but then put a framework around an agent much like we do a trustee where they are compelled to act in the best interests of the beneficiaries (the human that created them in this case).
Anyway it's got me thinking in a different way.
Asking it to draft was weakening my own skills.
An other thing for me is that it has gotten a lot harder for small teams with few ressources, let one person, to release anything that can really compete with anything the big player put out.
Quite a few years back I was working on word2vec models / embeddings. With enough time and limited ressources I was able to, through careful data collection and preparation, produce models that outperformed existing embeddings for our fairly generic data retrieval tasks. You could download from models Facebook (fasttext) or other models available through gensim and other tools, and they were often larger embeddings (eg 1000 vs 300 for mine), but they would really underperform. And when evaluating on general benchmarks, for what existed back then, we were basically equivalent to the best models in English and French, if not a little better at times. Similarly later some colleagues did a new architecture inspired by BeRT after it came out, that outperformed again any existing models we could find.
But these days I feel like there is nothing much I can do in NLP. Even to fine-tune or distillate the larger models, you need a very beefy setup.
I don't know how I'm burnt out from making this thing do work for me. But I am.
AI is the red herring that'll waste all our attention until it's too late.
And most of that new capacity will be natural gas. That increase would basically whipe out the reduction in CO2 emissions the USA has had since 2018.
> And most of that new capacity will be natural gas. That increase would basically whipe out the reduction in CO2 emissions the USA has had since 2018.
Emissions in 2018 were ~5250M metric ton and in 2024 it was 4750M. That is a reduction of 10% total emissions. Without going into calculations of green electricity and such, its still safe to say AI using 10% of the grid would not completely wipe out the reduction.
[0]: https://www.statista.com/statistics/183943/us-carbon-dioxide...
Transportation, especially ALL transportation, does a LOT. You're looking for ROI not the absolute values. I think it's undeniable that the positive economic effect of every car, truck, train, and plane is unfathomably huge. That's trains moving minerals, planes moving people, trucks transporting goods, and hundreds of combinations thereof, all interconnected. Literally no economic activity would happen without transportation, including the transition to green energy sources, of which would improve the emissions from transportation.
I think it might be more emissions-efficient at generating value than AI by a factor exceeding the 7.5x energy use. Moving rocks from (place with rocks) to (place that needs rocks) continues to be just an insanely good thing for humanity.
Also, I'm not sure about your math. 4% would be 4% of the whole like in a pie chart, not 4% of the remainder after removing one slice. 4% AI, 30% transportation, 66% other. I don't know where that 40% is from.
40% is for energy use in the US in the form of electricity. It was a rough number that I pulled from my memory. It is roughly right though. Check https://www.eia.gov/energyexplained/us-energy-facts/
AI is not currently 4% of the energy market of the US. Only the grid. I should have been more clear about the ALL ENERGY vs GRID distinction.
> I think it might be more emissions-efficient at generating value than AI by a factor exceeding the 7.5x energy use. Moving rocks from (place with rocks) to (place that needs rocks) continues to be just an insanely good thing for humanity.
I really made no statement on the value of doing things. Transportation is obviously very valuable. I just wanted a more fact based conversation.
I don't follow. The comparison is 30% of energy use for transportation vs 4% for AI, and soon 30% for transportation vs 10% for AI.
And that leaves a 6:1 ratio assuming projections run true. It very well might be possible to get efficiency wins from the transportation sector that outweigh growth in AI.
But yeah, there's way worse industries out there when it comes to climate change impact.
Ai buzz and now we are building giga factories. It stands for gigawatt usage, no less target.
It is, of course, because it barely uses any energy.
If you want to point at causes of climate change, look no further than adtech. It's the driving force behind our overconsumption.
And it has perhaps an even longer list of reasons to hate it.
So this is not a good reason to oppose AI. Now the sheer energy it requires does mean we might want to go nuclear though.
Natural gas is nice though because it does pollute the air far less than coal.
You might argue the EPA only repealed that because of political agendas, but the same argument could be made for why it was passed.
A lot of people got very rich off the fear mongering from climate alarmists.
And, you may be right, it may not be that big a deal and that we're being alarmists, but it seems like we currently have the tools to slow it down greatly. Why not be on the safe side and use them?
... but to be honest, guessing my opinion won't sway you in any way, still thought I'd try. thanks!
I'm finding the detractors worse than the hype, because it seems like a certain subset of detractors [0] formed their opinion on AI in late 2022/early 2023 when ChatGPT came out (REALLY!? Over 3 years ago!?) and then never updated their opinions since then. They'll say things like "why would I want to consume X amount of energy and Y amount of water just to get a wrong answer?"
In other words, the people who think generative AI is an absolutely worthless and useless product are more annoying than the ones that think it's going to solve all the world's problems. They have no idea how much AI has improved since it reached center stage 3 years ago. Hallucinations are exceptionally rare now, since they now rely on searching for answers rather than what was in its training data.
We got Claude Desktop at work and it's been a godsend. It works so much better to find information from Confluence and present it to me in a digestible format than having to search by hand and combing through a dozen irrelevant results to find the one bit of information I need.
[0] For the purpose of this comment, this subset is meant to be detraction based on the quality of the product, not the other criticisms like copyright/content theft concerns, water/energy usage, whether or not Sam Altman is a good person, etc.
Detractors, doomers, and techno-pessimists have got to be the most consistently wrong group in history. https://pessimistsarchive.org/
But I do think humanity is worse off because of it. So I'm a detractor in that way. :)
Well, I wouldn't go that far, but the hallucinations have moved up to being about more complicated things than they used to be.
Also, I've seen a few recent ones that "think" (for lack of a better word) that they know enough about politics to "know" they don't need to search for current events to, for example, answer a question about the consequences of the White House threatening military action to take Greenland. (The AI replied with something like "It is completely inconceivable that the US would ever do this").
This is such a perfect example of the mania behind this rollout.
There's no way you can make the financials work here compared to JetBrains spending the same millions spent on AI infrastructure and instead building better search in Confluence. Confluence search SUCKS, but that's just a lack of focus (or resources) on building a more complex, more robust solution. It's a wiki.
Either way, making a more robust search is a one time cost that benefits everyone. Instead, you're running a piece of software that goes directly to Anthropic's bank account, and to the data centers and to hyper scalers. Every single query must be re-run from scratch, costing your company a fortune, that if not managed properly will come out of spending that money elsewhere.
If I could pay a world class architect $1.50 to give me tips on how to maximize sunlight in my loft I would.
Would it be nice if confluence just had a robust search that had a one time cost and then benefited everyone thereafter? Sure, but that's not the current reality, and I do not have control over their actions. I can only control mine.
It seems ridiculous right now because we don’t have hardware to accelerate the LLMs, but in 5 years this will be trivial to run.
Even with an LLM agent getting cheaper to run in the future, it's still fundamentally non-deterministic so the ongoing cost for a single exploration query run can never get anywhere near as cheap as running a wiki with a proper search engine.
I mean, you can get mad at people you made up in your head, that's a thing people do, but this caricature falls in the same comforting bucket as "anyone who doesn't like <thing I like> is just ignorant/stupid" and "if you don't like me you're just jealous".
Maybe non-straw people have criticisms that aren't all butterflies and rainbows for good reasons, but you won't get to engage with them honestly and critically if you're telling yourself they're just ignorant from the start.
For example, I will bet that non-straw people will take issue with this, and for good reasons:
> Hallucinations are exceptionally rare now
In contrast, what harm do those detractors cause? They don't generate as much code per hour?
The "harm" (if you can call it that) is clear, detractors slow the pace of progress with meaningless and incorrect hand-wringing. A lack of progress harms everyone (as evidenced our amazing QoL today compared to any historical lens.)
Considering our climate, political and economic situation, I'd say not only is slowing the pace of progress not harmful, it's actually imperative for our long-term survival.
Also we need detractors because if we race into any technological advance too quickly we may cause unnecessary harm. Not all progress is without harms, and we need to be responsible about implementing it as risk-free as possible.
I honestly am finding Codex considerably better, as much as I despise OpenAI.
On the contrary. I update my opinion all the time, but every time I try the latest LLM it still sucks just as much. That is why it sounds like my opinion hasn't changed.
I say all of that to establish that I'm not a reflexive critic when I tell you, hallucinations are absolutely not exceptionally rare now. On multiple occasions this week (and it's only Tuesday!) I've had to disprove a LLM hallucination at work. They're just not as fun to talk about anymore, both because they're no longer new and because straightforward guardrails are effective at blocking the funny ones.
AI is alright. It's moderately useful, in certain contexts it speeds me up a lot, in other contexts not so much.
I also think that the economics of it make no sense and that it is, generally, a destructive technology. But it's not up to me to fix anything, I just try to keep on top of best practices while I need to pay bills.
The economics bit is not my problem though. If all AI companies go bust and AI services disappear I can 100% manage without it.
We're in "too big to fail" territory, if we handled the recession we were heading towards/in years ago, instead of letting AI hype distract and redirect massive amounts of investment, attention and labor from elsewhere, we might have been in a better position.
Though, my cynical take is that the investor class seemed dead-set on forcing us all to weave LLMs deep into our corporate infrastructures in a way that I'm not too sure it will ever "disappear" now. It'll cost just as much to detangle it as it was to adopt it.
The way we talk about "hallucinations" is extremely unproductive. Everything an LLM outputs is a hallucination. Just like how human perception is hallucination. These days I pretty much only hear this word come up among people that are ignorant of how LLMs work or what they're used for.
I've been asked why LLMs hallucinate. As if omniscient computer programs are some achievable goal and we just need to hammer out a few kinks to make our current crop of english-speaking computers perfect.
If AGI is born from these efforts, it will likely be controlled by people who stand to lose the most from solving those issues. If an OpenAI-built AGI told Sam Altman that reducing wealth inequality requires taxing his own wealth, would he actually accept that? Would systems like that get even close to being in charge?
All but one of them simultaneously, in fact. The one being left out: wanting to keep existing.
If you want to "keep existing" AGI happening is probably your only hope.
If you want to keep existing, slow down, make sure AGI is aligned first, and go into cryo if necessary.
If you don't want to keep existing, that doesn't mean you get to risk the rest of us.
How, exactly, does more and better tech help with the fundamentally sociological issues of power distribution, wealth inequality, surveillance, etc? Are you operating on the assumption that a machine superintelligence will ignore the selfish orders of whoever makes it and immediately work to establish post-scarcity luxury space communism?
So tired of seeing this trope. Data center energy expenditure is like less than 1% of worldwide energy expenditure[1]. Have you heard of mining? Or agriculture? Or cars/airplanes/ships? It's just factually wrong and alarmist to spread the fake news that AI has any measurable effect on climate change.
[1] https://www.iea.org/reports/energy-and-ai/energy-supply-for-...
https://www.selc.org/news/resistance-against-elon-musks-xai-...
Thats not accurate. The estimate is about 2 billion in 25 years.
https://www.who.int/news-room/fact-sheets/detail/ageing-and-...
We also have models for how that works at a country level because we have countries that have far exceeded that.
And the vast majority of 60 year olds are still self sufficient and economically productive.
Average global retirement age is around 65 and in most countries it’s creeping towards 70. And percent of world population over 70 looks much more manageable over the time span we can realistically model.
Some people have been concerned with this kind of politics all along. Some people are realizing they should be now, because of AI. And that's okay; both groups can still work together.
The most frustrating about AI I find is that it is (trying to) replacing things that I like; writing, art, programming while leaving me with the things I absolutely hate like testing, chores etc. I do like reading code, so thats not a big issue; I spent most of the day doing that before AI, however, being a fulltime QA was always my nightmare but here we are; AI sucks at it for a more than trivial frontend and backend its not that good at either (I should write the tests or rather tell the AI in detail what to test for otherwise it will just positively test what it indeed wrote; not what the spec says it had to write in many cases).
But no, I like talking about AI, just not so much about slop, trivial usage (show hn; here is a SaaS to turn you into a rabbit!) or hyperbole (our jubs!) (Although I do believe it is the end of code; like said; I read and review code all day but have not written much for the past 6 months while working on complex non trivial SaaS projects; I am with antirez; it is automatic coding, not vibecoding for me).
Everythink devolves from “cool that was a nice single video” to “here’s my schtick…. AGAIN”.
I also have to say that I don't use AI in my personal or professional life. And that is simply because I haven't felt any need to use it.
My technical interests are varied, and it's so boring to come to HN and see that a third (or more) of the front page is about AI.
Enough already. Let's talk about other things! And yes, I know, I should be a part of the solution and submit more articles.
I'm currently reading non-things by Byung-Chul Han, which is an interesting exploration of internet's impact on humanity/humans. Haven't finished it yet, but enjoying it so far.
Currently it feels a bit like everyone is talking about what new editor they're using. I don't care about that type of developer tooling very much. AI isn't coming up with some exciting new database, type system, etc etc.
"Look at how I'm able to web dev x% faster" because of LLMs is boring.
> seems to have devolved into three different people’s (almost identical) Claude code workflow
I do feel like I've seen a number of those articles.
I've been spamming some auto research loops, and it is so addictive. Think about how many of humanity's problems will be solved because of this. Of course, it will also disappoint, like, we are still waiting for flying cars, but man, this is a unique moment in history.
At least I'm not tired of talking about how it's killing websites and filling everything with spam. I have spent most of a decade building a useful resource, and Google AI overviews has killed my traffic. It killed everyone's traffic. This thing gave me purpose, and I'm watching AI slowly strangle it.
I mourn the death of the independent web, and it frightens me that this is still the happy stage. We haven't yet felt the effect of stiffing content creators, and the LLM tools haven't yet begun to enshittify.
I am tired of discussions about agentic coding, but I would feel a lot better if we acknowledged all the harm being caused. Big tech went all in on this, stealing everything, putting everyone out of work, using up all resources with no regards for consequences, and they threaten to kill the economy if we don't let them have their way.
I feel like we are heading for a much worse place as a society, and all we can talk about is how to 10x our bullshit jobs, because we're afraid of falling behind.
I’m just kidding. LinkedIn feed became so unbearable, that I had to install an extension to turn it off.
Large organizations are making major decisions on the basis of it. Startups new and old will live and die by the shift that it's creating (is SaaS dead? Well investors will make it so). Mass engineering layoffs could be inevitable.
Sure. I vibe coded a thing is getting pretty tired. The rest? If anything we're not talking about it enough.
I'm so exhausted by this and ready for the economic crash.
Sounds very much like this blog I read too… he laughs at AI in his workplace a lot Www.sometimesworking.com
I'm somewhat tired of seeing the same rehashed claims of future ability, non-ability, profit, loss.
I actually like talking about the implications, future risks and challenges of AI. I have made submissions on ways AI should be regulated to benefit society. The problem is the assumption of what is happening and what will happen.
To many people seem to enter the conversation feeling that the absence of doubt is the same thing as being informed.
And especially people making claims based on premises that they seem to believe that if they build big enough towers on them, they will become true.
The number one thing that bothers me in all this, is people assuming the contents of the minds of others.
I find the pathologising of Sam Altman to be the most egregious form of this. It is one thing to disagree with someone's decisions, another thing to disagree with their stated opinions, but to decide upon a person's character based upon what you believe they are thinking in their private thoughts is simply projection.
I know this is an opinion of little worth to many, but my impression of Sam Altman is just a person who has different perspectives to me. The capitalist tech world he lives in would inevitably shape different values to me. What I have seen of him is consistent with a sincere expression of values. I can accept that a person might do something different to what I would, even the opposite of what I want while believing that they can be doing so for reasons that seem to be morally the right thing to do.
This also happened with cryptocurrency. Crypto advocates believe that it is a good thing for the world. Too many consider those who believe that crypto could benefit society to be evil. There is a difference between being wrong and being evil. No matter how certain you are you can still be wrong, in fact beyond a point I would say increased certancy would indicate a higher likelihood of being wrong.
So I'm happy to talk about AI. I have plenty to learn. I wonder if others went in with the goal to learn whether they would find it less tiring.
It's worse when there's a colleague of yours encouraging that by using AI blindly, piling up technical debt just to move at the pace that Management expects after signing you all up on some AI tool.
At the end of the day, everyone is talking about AI. For AI or against AI, it doesn't really matter.
The analogy is someone from the 19th century talking about their slaves all day which is of course nonsense because they had other things to talk about.
Bored of hearing about it, bored of reading about it.
I love using these LLM tools, but honestly, it feels like every man and his dog has something to say about it, and is angling to make a quick buck or two from it.
And the slop, oh my goodness, it's never-ending on every site and service.
Tack on to that the increasing number of political stuff on here as well just makes it less and less an interesting place to visit.
Don't agree with the angry mob on the political stuff especially and you get downvoted/flagged into oblivion.
Just another echo chamber looking to have viewpoints confirmed in yet another one of the disappearing places online that foster any level of intellectual curiosity.
Never thought I'd feel nostalgic about that era...
AI is especially sensitive to this. Unlike coding, where giving away the secret sauce also makes you look smart, divulging AI secrets only demystifies you -- revealing the shriveling man behind the Wizards curtain.
So anyone boasting about AI is likely not doing anything useful with it.
Similar to finance tips, btw.
Or another thought; why is it that a stochastic parrot can solve logic puzzles consistently and accurately? It might not be 100%, but it’s still much better than what you might expect from a markov model of ngrams.
Openclaw is only sort of interesting. How to vibe code your first product is uninteresting. Claims about productivity increase from model usage are speculative and uninteresting. Endless think pieces on the effects of AI slop are uninteresting. There’s a lot of hype and grift and bullshit that is downstream of this very interesting technology, and basically none of that is interesting. The cool parts are when you actually open the models up and try to figure out what’s going on.
So no, I’m not bored of talking about AI. I’m not sure I ever will be. My suspicion is that those who are bored of it aren’t digging deep enough. With that said, that will likely only be interesting to people who think math is fun and cool. On the whole, AI is unlikely to affect our lives in proportion to the ink spilled by influencers.
The greats on youtube are also worth watching: 3B1B, numberphile, etc.
[0] https://youtube.com/playlist?list=PLAqhIrjkxbuWI23v9cThsA9Gv... [1] https://youtu.be/qx7hirqgfuU?si=8zmrbazuvnz379gk
The short answer, as far as I’m aware, is that no one really knows. The longer answer is that we have a lot of partial answers that, in my mind, basically boil down to: model architectures draw a walk through the high dimensional vector space of concepts, and we’ve tuned them to land on the right answer. The fact that they do so consistently says something about how we encode logic in language and the effectiveness of these embedding/latent spaces.
It can't. As you say in the very next sentence. If it isn't solving any given puzzle with a 100% success rate, but randomly failing, then it isn't consistent.
As bad as the AI hype wave is now, I can't help but wonder if it could have been even worse.
There are other interesting things in the world today, and HN is overwhelmed with pretend intelligence.
Hype, detractors, ALL OF IT!
Maybe a separate web page or RSS feed could be created that is dedicated to the subject...
Then we can get back to the unglamorous, boring, thankless task of delivering business value to paying clients, and the public discourse will no longer be polluted by our inane witterings.
I used to have this idea that if I built something cool it would be valuable to donate it to the world for free. But now increasingly I'd be just making a donation to the training data, and on top of this I'm in competition with AI slop. Most people won't tell the difference and won't care. The noise floor for doing absolutely anything collaboratively on the computer is now 10x higher than it was before, and I'm basically checked out at this point. Even HN is becoming tiring to read since I think around 10-15% of comments that I read are AI generated. When that number reaches 30% I'm done forever, gone. My life is too short to waste time on this shit.
You could use AI to do it! Fight fire with fire.
I'm neutral on AI - so far it seems useful but flawed. But I don't want to hear about it constantly.
If you're reading this and your life hasn't been thrown into disarray you're likely just behind the times. There are a lot of people who are deep in tech who still don't understand what agents and LLM's can do
I'd love for discussions of the tech to stop with the genAI version of the cryptobro cry "have fun being poor". It's mildly insulting and adds literally nothing to the conversations.
(Not meaning to single you out, just using it as an example. This is a very common rhetorical problem with most of the evangelism.)
The detractors are so bizarre to me. I think it's because I work at a big tech that has so thoroughly wired AI into everything we do, and the benefits are so undeniable and totally perspective changing, that it's like arguing with someone that thinks the sun revolves around the earth.
So if you aren't doing something cool with AI, it's probably because you aren't empowered to at your company, or because you simply aren't taking the initiative. Seems like a pretty even split on HN.
and why do so many articles or comments have a general approach of 'It's great and if you don't think it is it's because you don't understand it.'
"Yes" Proceeds to talk about AI.
> At serious risk of sounding like a heretic here, but I’m kinda bored of talking about AI.
Umm.
> I get it, AI is incredible. I use it every day, it’s completely changed my workflow. I recently started a new role in a tricky domain working at web scale (hey, remember web scale?) and it’s allowed me to go from 0-1 in terms of productivity in a matter of weeks.
It’s all positives. So what’s the problem?
There isn’t a problem with AI. Of course. It’s just the discourse around it is “boring”. And the managers are lame about it.
And what has been the AI discourse for the last few years. The same formula.
- AI is either good
- ... or it is the best thing to have happened to Me
- But I have feelings[1] or concerns about everything around AI, like the discourse, or people having two-hundred concurrent AI agents mania
It’s all just grease for the AI Inevitabilism bytemill.
[1] https://news.ycombinator.com/item?id=47487774
> … And yes, I’m painfully aware of the irony of a post about moaning about posts about AI. Sorry.
OP can’t even resign himself to being a Type. Sigh. “I know what I just did hehe”
Very self-aware.
And now 117 points and 53 comments in 23 minutes.
> And this one will be different? I think you're talking about my blog post here, in which case no, I'm afraid not. Hence the admission at the bottom.
>Umm. ??
> It’s all positives. So what’s the problem? The article is trying to say that these things are great, but the level of conversation leads to a lack of novelty.
> It’s just the discourse around it is “boring”. And the managers are lame about it. Exactly.
> OP can’t even resign himself to being a Type. Sigh. “I know what I just did hehe” Very self-aware.
Is this sarcasm?
Is anybody else bored of talking about AI? I’m beyond bored.
Why on earth is the parent comment downvoted? the title of the TFA asks a question. This statement directly answers that question. Seems very on-topic.