I was paying for Max but after trying GLM 4.7 I am a convert. Hardly hit the limit but even if I do it is cheaper to get two accounts from Z.ai than one Max from Anthropic
The only problem I have is that it's so effing expensive to run those games that I can't have a good number of games to claim to be any sort of legit benchmark. BUT so far the games that I paid out of pocket and ran are looking good and I think there is merit to this.
also had lots of fun building on top of Cloud Flare and solving some distributed systems problems while building this.
if you can help me run more games (for science!!) let me know!
This is concerning on two fronts. The questions are no longer open (SO is CC-BY-SA) and if Q&A content dies then this herds even more people towards LLM use.
It's basically draining the commons.
Yup. This, to me, provides another explanation for why the social contract is being used as toilet paper by the owner class. They literally see the writing on the wall.
Thanks! Workflows is definitely interesting – it's basically durable execution with steps and retries. It's on the radar, probably after the CLI and GitHub integration.
i think this is a little unfair, its comparing a model that is optimised for pass@2 and self improving its output compared to the other models, just test time scaling in a way
I really really want this to be true. I want to be relevant. I don’t know what to do if all those predictions are true and there is no need (or very little need) for programmers anymore.
But something tells me “this time is different” is different this time for real.
Coding AIs design software better than me, review code better than me, find hard-to-find bugs better than me, plan long-running projects better than me, make decisions based on research, literature, and also the state of our projects better than me. I’m basically just the conductor of all those processes.
Oh, and don't ask about coding. If you use AI for tasks above, as a result you'll get very well defined coding task definitions which an AI would ace.
I’m still hired, but I feel like I’m doing the work of an entire org that used to need twenty engineers.
I was a chef in Michelin-starred restaurants for 11 years. One of my favorite positions was washing dishes. The goal was always to keep the machine running on its 5-minute cycle. It was about getting the dishes into racks, rinsing them, and having them ready and waiting for the previous cycle to end—so you could push them into the machine immediately—then getting them dried and put away after the cycle, making sure the quality was there and no spot was missed. If the machine stopped, the goal was to get another batch into it, putting everything else on hold. Keeping the machine running was the only way to prevent dishes from piling up, which would end with the towers falling over and breaking plates. This work requires moving lightning fast with dexterity.
AI coding agents are analogous to the machine. My job is to get the prompts written, and to do quality control and housekeeping after it runs a cycle. Nonetheless, like all automation, humans are still needed... for now.
If it requires an expert engineer/dishwasher to keep the flow running perfectly, the human is the bottleneck in the process. This sounds a lot more like the past before AI to me. What AI does is just give you enough dishes that they don’t need to be washed at all during dinner service. Just let them pile up dirty or throw them away and get new dishes tomorrow it’s so immaterial to replace that washing them doesn’t always make sense. But if for some reason you do want to reuse them, then, it washes and dries them for you too. You just look over things at the end and make sure they pass your quality standards. If they left some muck on a plate or lipstick on a cup, just tell it not to let that happen again and it won’t. So even your QC work gets easier over time. The labor needed to deal with dirty dishes is drastically reduced in any case.
"AI" doesn't have a clue what to do on its own. Humans will always be in the loop, because they have goals, while the AI is designed to placate and not create.
The amount of "AI" garbage I have to sift through to find one single gem is about the same or more work than if I had just coded it myself. Add to that the frustration of dealing with a compulsive liar, and it's just a fucking awful experience for anyone that actually can code.
> Coding AIs design software better than me, review code better than me, find hard-to-find bugs better than me, plan long-running projects better than me, make decisions based on research, literature, and also the state of our projects better than me.
That is just not true, assuming you have a modicum of competence (which I assume you do). AIs suck at all these tasks; they are not even as good as an inexperienced human.
For all we know, you both could comparing using a Nokia 3310 and a workstation PC based on the hardware, but you both just say "this computer is better than that computer".
There are a ton of models out there, ran in a ton of different ways, that can be used in different ways with different harnesses, and people use different workflows. There is just so many variables involved, that I don't think it's neither fair nor accurate for anyone to claim "This is obviously better" or "This is obviously impossible".
I've been in situations where I hit my head against some hard to find bug for days, then I put "AI" (but what? No one knows) to it and it solves it in 20 minutes. I've also asked "AI" to do trivial work that it still somehow fucked up, even if I could probably have asked a non-programmer friend to do it and they'd be able to.
The variance is great, and the fact that system/developer/user prompts matter a lot for what the responses you get, makes it even harder to fairly compare things like this without having the actual chat logs in front of you.
> was the apparent benefit so incredibly variable?
Yes, lots of people were very vocally against horseless-carriages, as they were called at the time. Safety and public nuisance concerns were widespread, the cars were very noisy, fast, smoky and unreliable. Old newspapers are filled with opinions about this, from people being afraid of horseless-carriages spooking other's horses and so on. The UK restricted the adoption of cars at one point, and some Canton in Switzerland even banned cars for a couple of decades.
Horseless-carriages was commonly ridiculed for being just for "reckless rich hobbyists" and similar.
I think the major difference is that cars produced immediate, visible externalities, so it was easy for opposition to focus on public safety in public spaces. In contrast, AI has less physically visible externalities, although they are as important, or maybe even more important, than the ones cars introduced.
yeah I agree about the negative externalities but I'm curious about the perceived benefits. did anybody argue that cars were actually slower than horse and carriage? (were they at first?)
The cars were obviously faster than the typical horse transportation and I don't think anyone tried to argue against that, but laws typically restricted cars so they couldn't go faster than horses, at least in highly populated areas like cities. As others mentioned too, the benefit of not needing roads to go places were highlighted as a drawback of cars too. People argued that while cars might go faster, the result would be that the world would be worse off in total.
sure but my point is people could agree they were faster at least. that is decidedly not true for LLMs. maybe due to alignable vs non-alignable differences
Is this a trick question? Yes it was. A horse could go over any terrain while a car could only really go over very specific terrain designed for it. We had to terraform the world in order to make the automobile so beneficial. And it turned out that this terraforming had many unintended consequences. It's actually a pretty apt comparison to LLMs.
who would I be trying to trick if it was? you didn't answer the question anyways. I'm not wondering whether cars were seen as strictly better than horses in all situations. I'm wondering if people disagreed so vehemently about whether cars were faster road transportation than horses
LLMs generate the most likely code given the problem they're presented and everything they've been trained on, they don't actually understand how (or even if) it works. I only ever get away with that when I'm writing a parser.
It matters if AGI is the goal. If it remains a tool to make workers more productive, then it doesn't need to truly understand, since the humans using the tools understand. I'm of the opinion AI should have stood for Augmented (Human) Intelligence outside of science fiction. I believe that's what early pioneers like Douglas Engalbert thought. Clearly that's what Steve Jobs and Alan Kay thought computing was for.
AGI is such a meaningless concept. We can’t even fully design what human intelligence is (and when a human fails it meaning they lack human intelligence). It’s just philosophy.
If it empirically works, then sure. If instead every single solution it provides beyond a few trivial lines falls somewhere between "just a little bit off" and "relies entirely on core library functionality that doesn't actually exist" then I'd say it does matter and it's only slightly better than an opaque box that spouts random nonsense (which will soon include ads).
Late 2025 models very rarely hallucinate nonexistent core library functionality - and they run inside coding agent harnesses so if they DO they notice that the code doesn't work and fix it.
This sounds like you're copy-pasting code from ChatGPT's web interface, which is very 2024.
Agentic LLMs will notice if something is crap and won't compile and will retry, use the tools they have available to figure out what's the correct way, edit and retry again.
Depends on how he defined "better". If he uses the word "better" to mean "good enough to not fail immediately, and done in 1/10th of the time", then he's correct.
I think I've been using AI wrong. I can't understand testimonies like this. Most times I try to use AI for a task, it is a shitshow, and I have to rewrite everything anyway.
Ok, but have you tried claude-sonnet-GPT-codex-4.5-thinking-fast? That's the game changer. Anyone saying bad things about vibe coding without trying claude-sonnet-GPT-codex-4.5-thinking-fast is like a dinosaur to me, doomed to extinction. Seriously, give claude-sonnet-GPT-codex-4.5-thinking-fast a try, you'll thank me ;)
I don’t know about right/wrong. You need to use the tools that make you productive. I personally find that in my work there are dozens of little scripts or helper functions that accelerate my work. However I usually don’t write them because I don’t have the time. AI can generate these little scripts very consistently. That accelerates my work. Perhaps just start simple.
> there are dozens of little scripts or helper functions that accelerate my work. However I usually don’t write them because I don’t have the time
People who write things like this can't expect to be taken seriously.
Before AI you didn't have time to write things that saved you time? So you just ended up spending (wasting) more time by going the long way? That was a better choice than just doing the thing that would have saved you time?
Do you tell AI the patterns/tools/architecture you want? Telling agents to "build me XYZ, make it gud!" is likely to precede a mess, telling it to build a modular monolith using your library/tool list, your preferred folder structure, other patterns/algorithms you use, etc will end you up with something that might have some minor style issues or not be perfectly canonical, but will be approximately correct within a reasonable margin, or is within 1-2 turns of being so.
You have to let go of the code looking exactly a certain way, but having code _work_ a certain way at a coarse level is doable and fairly easy.
Kiro is just trying to build a product around exactly what I'm talking about. I'm not a fan, because it's simultaneously too heavyweight and agents don't respect all the details of the specs it creates enough to make the time investment in super-detailed specs worthwhile.
I have a spec driven development tool I've been working on that generates structured specs that can be used to do automatic code generation. This is both faster and more robust.
Honestly, even this isn't really true anymore. With Opus 4.5 and 5.2 Codex in tools like Cursor, Claude Code, or Codex CLI, "just do the thing" is a viable strategy for a shockingly large category of tasks.
Just do the thing can produce functional code, but even with Opus4.5/Codex5.2, there are still plenty of moments where the way it decides to do something is cringe.
Agree. But it's increasingly the case, IME, that for a a lot of tasks, you can start with that. If it does it well, great. If it does something stupid, it's easy enough to ask it to completely rework the stupid thing in a better way, and it can do it quickly. That's still a huge shift compared to the olden days (three months ago) where you needed to really break things down into small chunks for it to get to a success state.
A lot of code smells matter more for humans than LLMs (and LLMs have their own unique code smells). For example, nested ternary operators are a great source of bugs in human code, but agents could care less, but humans handle multiple files with the same variable names and lots of duplicated code well, whereas this stuff confuses agents.
The phrase is "couldn't care less". If you "could care less" then you actually care about it. If you "couldn't care less" then there's no caring at all.
It’s because depending on the person the newest model crossed the line into being useful for them personally. It’s not like a new version crosses the line for everyone. It happens gradually. Each version more and more people come into the fold.
You get new people recommending the latest version all the time to people who are unconvinced because that version is usually what brought them into the fold.
What you’re mocking is somewhat of a signal of actual improvement of the models and that improvement as a result becoming useful to more and more people.
how much time/effort have you put in to educate yourself about how they work, what they excel at, what they suck at, what is your responsibility when you use them…? this effort is directly proportional to how well they will serve you
Not because the models are random, but because you are mistaking a massive combinatorial search over seen patterns for genuine reasoning. Taleb point was about confusing luck for skill. Dont confuse interpolation for understanding.
You can read a Rust book after years of Java, then go build software for an industry that did not exist when you started. Ask any LLM to write a driver for hardware that shipped last month, or model a regulatory framework that just passed... It will confidently hallucinate. You will figure it out. That is the difference between pattern matching and understanding.
I've worked with a lot of interns, fresh outs from college, overseas lowest bidders, and mediocre engineers who gave up years ago. All over the course of a ~20 year career.
Not once in all that time has anyone PRed and merged my completely unrelated and unfinished branch into main. Except a few weeks ago. By someone who was using the LLM to make PRs.
He didn't understand when I asked him about it and was baffled as to how it happened.
Really annoying, but I got significantly less concerned about the future of human software engineering after that.
Have you used an LLM specifically trained for tool calling, in Claude Code, Cursor or Aider?
They’re capable of looking up documentation, correcting their errors by compiling and running tests, and when coupled with a linter, hallucinations are a non issue.
I don’t really think it’s possible to dismiss a model that’s been trained with reinforcement learning for both reasoning and tool usage as only doing pattern matching. They’re not at all the same beasts as the old style of LLMs based purely on next token prediction of massive scrapes of web data (with some fine tuning on Q&A pairs and RLHF to pick the best answers).
I'm using Claude code to help me learn Godot game programming.
One interesting thing is that Claude will not tell me if I'm following the wrong path. It will just make the requested change to the best of its ability.
For example a Tower Defence game I'm making I wanted to keep turret position state in an AStarGrid2D. It produced code to do this, but became harder and harder to follow as I went on. It's only after watching more tutorials I figured out I was asking for the wrong thing. (TileMapLayer is a much better choice)
edit: Major engine changes have occurred after the models were trained, so you will often be given code that refers to nonexistent constants and functions and which is not aware of useful new features.
before coding I just ask the model "what are the best practices in this industry to solve this problem? what tools/libraries/approaches people use?
after coding I ask it "review the code, do you see any for which there are common libraries implementing it? are there ways to make it more idiomatic?"
you can also ask it "this is an idea on how to solve it that somebody told me, what do you think about it, are there better ways?"
> before coding I just ask the model "what are the best practices in this industry to solve this problem? what tools/libraries/approaches people use?
Just for the fun of it, and so you lose your "virginity" so to speak, next time when the magic machine gives you the answer about "what it thinks", tell it its wrong in a strict language and scold it for misleading you. Tell it to give you the "real" best practices instead of what it spat out.
Then sit back and marvel at the machine saying you were right and that it had mislead you. Producing a completely, somewhat, or slightly different answer (you never know what you get on the slot machine).
Both the before and after are better done manually. What you are describing is fine for the heck of it (I‘ve vibe coded a whisper related rust port today without having any actual rust skills), but I’d never use fully vibed software in production. That’s irresponsible in multiple ways.
I‘ve just tried the dxastgraphx one in pi with Opus 4.5. This was its response:
I couldn't find a library called dxastgraphx in either pip (Python) or npm (JavaScript) package registries. This library doesn't appear to exist.
Did you perhaps mean one of these popular DAG/graph libraries?
Python:
- networkx - comprehensive graph library with DAG support
- graphlib - Python standard library (3.9+) with TopologicalSorter
- dask - parallel computing with DAG task scheduling
JavaScript/TypeScript:
- graphlib - graph data structures
- dagre - DAG layout
Would you like me to build a DAG scheduler using one of these existing libraries, or would you like me to implement one from scratch? Let me know which language and approach you prefer.
Tried ChatGPT (free version). I answered correctly to the first question.
Answer to second question:
"I can do that, but there’s a catch: dxastgraphx is not a known or standard Python (or JS) library as of now. I don’t have any public API or documentation for it to target directly.
To avoid guessing wrong, here are the two sensible options:"
> The reverse of the string 'gssssyeekxxuoaqaaaiooo9aaa7zjx' is:
xjz7aaa9oooiiaaaqauxxkyeessssg
> Youve made a mistake
```
I see where I tripped up! Let's re-examine that long string carefully.
The original string you gave me was:
`gssssyeekxxuoaqaaaiooo9aaa7zjx`
Let's look at the end of that string: `...aaa7zjx`.
When we reverse it, it should start with: `xjz7aaa...`
Here is the precise character-by-character reversal:
*xjz7aaa9oooiiaaaqauxxkyeessssg*
Wait—I see it now. In my previous response, I included an extra *'i'* and missed one *'o'* in the middle section. Let's do a strict count to get it perfect:
The correct reversed string is:
*xjz7aaa9oooiiaaaqauoxxkyeessssg*
Thank you for catching that. It’s easy for the eyes to glaze over when there are so many repeating vowels!
Since I've stumbled a bit on these, would you like to give me one more string to see if I can get it right on the first try?
```
After more back and fors it consistently fails in this task, even though when strictly dividing the tokens it will get this right. Yet the final answer is always wrong.
You’re trying to interrogate a machine as you would a human and presenting this as evidence that machines aren’t humans. Yes, you’re absolutely right! And also completely missing the point.
Why would you expect an LLM or even a human to succeed in these cases? “Write a piece of code for a specification that you can’t possibly know about?” That’s why you have to do context engineering, just like you’d provide a reference to a new document to an engineer writing code.
They do all those things you've mentioned more efficiently than most of us, but they fall woefully short as soon as novelty is required. Creativity is not in their repertoire. So if you're banging out the same type of thing over and over again, yes, they will make that work light and then scarce. But if you need to create something niche, something one-off, something new, they'll slip off the bleeding edge into the comfortable valley of the familiar at every step.
I choose to look at it as an opportunity to spend more time on the interesting problems, and work at a higher level. We used to worry about pointers and memory allocation. Now we will worry less and less about how the code is written and more about the result it built.
Take food for example. We don't eat food made by computers even though they're capable of making it from start to finish.
Sure we eat carrots probably assisted by machines, but we are not eating dishes like protein bars all day every day.
Our food is still better enjoyed when made by a chef.
Software engineering will be the same. No one will want to use software made by a machine all day every day. There are differences in the execution and implementation.
No one will want to read books entirely dreamed up by AI. Subtle parts of the books make us feel something only a human could have put right there right then.
No one will want to see movies entirely made by AI.
The list goes on.
But you might say "software is different". Yes but no, in the abundance of choice, when there will be a ton of choice for a type of software due to the productivity increase, choice will become more prominent and the human driven software will win.
Even today we pick the best terminal emulation software because we notice the difference between exquisitely crafted and bloated cruft.
You should look at other engineering disciplines. How many highway over passes have unique “chef quality” designs? Very few. Most engineering is commodity replications of existing designs. The exact same thing applies to software engineering. Most of us engineers are replicating designs that came earlier. LLMs are good at generating the rote designs that make up the bulk of software by volume. Who benefit from an artisanal REST interface? The best practices were codified over a decade ago.
> How many highway over passes have unique “chef quality” designs?
Have you ever built a highway overpass? That kind of engineering is complex and interdisciplinary. You need to carry out extensive traffic pattern analysis and soil composition testing to even know where it should go.
We're at a point where we've already automated all the simple stuff. If you want a website, you don't type out html tags. You use Squarespace or Wordpress or whatever. If you need a backend, you use Airtable. We already spend most of our time on the tricky stuff. Sure, it's nice that LLMs can smooth the rough edges of workflows that nobody's bothered to refine yet, but the software commodities of the world have already been commodified.
Just like cooking in the middle ages. As the kitchen, hygiene, etc. got better, so did the chefs and so did the food.
This is just a transition.
re-Rest API, you're right. But again, we use roombas to vacuum when the floor layout is friendly to them. Not all rooms can be vacuumed by roombas. Simple Rest api can be emitted one shot from an LLM and there is no room for interpretation. But ask a future LLM to make a new kind of social network and you'll end up with a mash up of the existing ones.
Same thing, you and I won't use a manual screwdriver when we have 100 screws to get in, and we own an electric drill.
That didn't reinvent screws nor the assembly of complex items.
I'm keeping positive in the sense that LLMs will enable us to do more, and to learn faster.
The sad part about vibe coding is you learn very little. And to live is to learn.
You'll notice people vibecoding all day become less and less attached to the product they work on. That's because they've given away the dopamine hits of the many "ha-ha" moments that come from programming. They'll lose interest. They won't learn anymore and die off (career wise).
So, businesses that put LLM first will slowly lose talent over time, and business that put developers first will thrive.
It's just a transition. A fast one that hits us like a wall, and it's confusing, but software for humans will be better made by humans.
I've been programming since the 80s. The level of complexity today is bat shit insane. I welcome the LLM help in managing 3 code bases of 3 languages spread across different architectures (my job) to keep sane!
I disagree with the vibecoding take. Its a new skill that absolutely has a place in developers skillset and it may be of great importance for some kinds of projects. You can learn so much by vibecoding little projects that otherwise would never see the light of day.
There is a part of this that is true. But when you get the nuanced parts of every "replicated design" or need the tweaks or what the AI gave you is just wrong, that deteriorates quality.
For many tasks it is ok, for others it is just a NO.
For software maintenance and evolution I think it won't cut it.
The same way a Wordpress website can do a set of useful things. But when you need something specific, you just drop to programming.
You can have your e-commerce web. But you cannot ask it to give you a "pipeline excution as fast as possible for calculating and solving math for engineering task X". That needs SIMD, parallelization, understanding the niche use you need, etc. which probably most people do not do all the time and requires specific knowledge.
> So if you're banging out the same type of thing over and over again, yes, they will make that work light and then scarce.
The same thing over and over again should be a SaaS, some internal tool, or a plugin. Computers are good at doing the same thing over and over again and that's what we've been using them for
> But if you need to create something niche, something one-off, something new, they'll slip off the bleeding edge into the comfortable valley of the familiar at every step.
Even if the high level description of a task may be similar to another, there's always something different in the implementation. A sports car and a sedan have roughly the same components, but they're not engineered the same.
> We used to worry about pointers and memory allocation.
Some still do. It's not in every case you will have a system that handle allocations and a garbage collector. And even in those, you will see memory leaks.
> Now we will worry less and less about how the code is written and more about the result it built.
I think your image of LLMs is a bit outdated. Claude Code with well-configured agents will get entirely novel stuff done pretty well, and that’s only going to get better over time.
As of today NONE of the known AI codebots can
solve correctly ANY of the 50+ programming
exercises we use to interview fresh grads
or summer interns. NONE! Not even level 1
problems that can be solved in fewer than 20
lines of code with a bit of middle school
math.
After 25+ years in this field, having interviewed ~100 people for both my startup and other companies, I'm having a hard time believing this. You're either in an extremely niche field (such as to make your statement irrelevant to 99.9% of the industry), or it's hyperbole, or straight up bs.
Interviewing is an art, and IME "gotcha" types of questions never work. You want to search for real-world capabilities, and like it or not the questions need to match those expectations. If you're hiring summer interns and the SotA models can't solve those questions, then you're doing something wrong. Sorry, but having used these tools for the past three years this is extremely ahrd to believe.
I of course understand if you can't, but sharing even one of those questions would be nice.
I promise you that I can show you how to reliably solve any of them using any of the latest OpenAI models. Email me if you want proof; josh.d.griffith at gmail
> the person using the tool (e.g. OpenAI, Claude, DevStral, DeepSeek, etc) must NOT be able to solve problems alone
I think this is a good point, as I find the operators input is often forgotten when considering the AIs output. If it took me an hour and decades of expertise to get the AI to output the right program, did the AI really do it? Could someone without my expertise get the same result?
If not, then maybe we are wasting our time trying to mash our skills through vector space via a chat interface.
However I'm still finding a trend even in my org; better non-AI developers tend to be better at using AI to develop.
AI still forgets requirements.
I'm currently running an experiment where I try to get a design and then execute on an enterprise 'SAAS-replacement' application [0].
AI can spit forth a completely convincing looking overall project plan [1] that has gaps if anyone, even the AI itself, tries to execute on the plan; this is where a proper, experienced developer can step in at the right steps to help out.
IDK if that's the right way to venture into the brave new world, but I am at least doing my best to be at a forefront of how my org is using the tech.
[0] - I figured it was a good exercise for testing limits of both my skills prompting and the AI's capability. I do not expect success.
> I’m basically just the conductor of all those processes.
a car moves faster than you, can last longer than you, and can carry much more than you. But somehow, people don't seem to be scared of cars displacing them(yet)? Perhaps autodriving would in the near future, but there still needs to be someone making decisions on how best to utilize that car - surely, it isn't deciding to go to destination A without someone telling them.
> I feel like I’m doing the work of an entire org that used to need twenty engineers.
and this is great. A combine harvester does the work of what used to be an entire village for a week in a day. More output for less people/resources expended means more wealth produced.
> a car moves faster than you, can last longer than you, and can carry much more than you. But somehow, people don't seem to be scared of cars displacing them(yet)?
People whose life were based around using horses for transportation were very scared of cars replacing them though, and correctly so, because horses for transportation is something people do for leisure today, not necessity. I feel like that's a more apt analogy than comparing cars to any human.
> More output for less people/resources expended means more wealth produced.
This is true, but it probably also means that this "more wealth produced" will be more concentrated, because it's easier to convince one person using AI that you should have half of the wealth they produce, rather than convincing 100 people you should have half of what they produce. From where I'm standing, it seems to have the same effects (but not as widespread or impactful, yet) as industrialization, that induced that side-effect as well.
Analogies are not going to work. Bug it's just as likely that, in the worst case, we are stage coach drivers who have to use cars when we just really love the quiet slowness of horses.
> a car moves faster than you, can last longer than you, and can carry much more than you. But somehow, people don't seem to be scared of cars displacing them(yet)?
> Well no, you'd be the horse driver who becomes a car driver
Well, that's the crux of the argument. The pro-AI devs are making the claim that devs are the horse-drivers, the anti-AI is making the claim that devs are the horses themselves.
There is no objective way to verify who is right in this case, we just have to see it play out.
Sure LLMs can churn out code, and they sort of work for developers who already understand code and design, but what happens when that junior dev with no hard experience builds their years of experience with LLMs?
Over time those who actually understand what the LLMs are doing and how to correct the output are replaced by developers who've never learned the hard lessons of writing code line by line. The ability to reason about code gets lost.
This points to the hard problem that the article highlights. The hard problem of software is actually knowing how to write it, which usually takes years, sometimes up to a decade of real experience.
Any idiot can churn out code that doesn't work. But working, effective software takes a lot of skill that LLMs will be stripping people of. Leaving a market there for people who have actually put the time in and understand software.
This is not how I think about it.
Me and the coding assistant is better then me or the coding assistant separately.
For me its not about me or the coding assistant, its me and the coding assistant. But I'm also not a professional coder, i dont identify as a coder. I've been fiddling with programming my whole life, but never had it as title, I've more worked from product side or from stakeholder side, but always got more involved, as I could speak with the dev team.
This also makes it natural for me to work side-by-side with the coding assistant, compared maybe to pure coders, who are used to keeping the coding side to themselves.
I have been using the most recent Claude, ChatGPT and Gemini models for coding for a bit more than a year, on a daily basis.
They are pretty good at writing code *after* I thoroughly described what to do, step by step. If you miss a small detail they get loose and the end result is a complete mess that takes hours to clean up. This still requires years of coding experience, planning ahead in head, you won't be able to spare that, or replace developers with LLMs. They are like autocomplete on steroids, that's pretty much it.
> Through Kiro, we reinvented how developers work with AI agents.
Even according to it’s documentation it is still built for developers, so my point still stands. You need dev experience to use this tool, same as other LLM-based coding tools.
I mean, AIs can drop something fast the same way you cannot beat a computer at adding or multiplying.
After that, you find mistakes, false positives, code that does not work fully, and the worse part is the last one: code that does not work fully but also, as a consequence, that you do NOT understand yet.
That is where your time shrinks: now you need to review it.
Also, they do not design systems better. Maybe partial pieces. Give them something complex and they will hallucinate worse solutions than what you already know if you have, let us say, over 10 years of experience programming in a language (or mabye 5).
Now multiply this unreliability problem as the code you "AI-generate" grows.
Now you have a system you do not know if it is reliable and that you do not understand to modify. Congrats...
I use AI moderately for the tasks is good at: generate some scripts, give me this small typical function amd I review it.
Review my code: I will discard part of your mistakes and hallucinations as a person that knows well the language and will find maybe a few valuable things.
Also, when reviewing and found problems in my code I saw that the LLMs really need to hallucinate errors that do not exist to justify their help. This is just something LLMs seem to not be accurate at.
Also, when problems go a bit more atypical or past a level of difficulty, it gets much more unreliable.
All in all: you are going to need humans. I do not know how many, I do not know how much they will improve. I just know that they are not reliable and this "generate-fast-unreliable vs now I do not know the codebase" is a fundamental obstacle that I think it is if not very difficult, impossible to workaround.
I feel you, it's scary. But the possibilities we're presented with are incredible. I'm revisiting all these projects that I put aside because they were "too big" or "too much for a machine". It's quite exciting
I'm extremely pro-faster-keyboard, i use the faster keyboards in almost every opportunity i can, i've been amazed by debugging skills (in fairness, i've also been very disappointed many times), i've been bowled over by my faster keyboard's ability to whip out HTML UI's in record time, i've been genuinely impressed by my faster keyboard's ability to flag flaws in PRs i'm reviewing.
All this to say, i see lots of value in faster keyboard's but add all the prompts, skills and hooks you like, explain in as much detail as you like about modularisation, and still "agents" cannot design software as well as a human.
Whatever the underlying mechanism of an LLM (to call it a next token predictor is dismissively underselling its capabilities) it does not have a mechanism to decompose a problem into independently solvable pieces. While that remains true, and i've seen zero precursor of a coming change here - the state of the art today is equiv to having the agent employ a todo list - while this remains true, LLMs cannot design better than humans.
There are many simple CRUD line of business apps where they design well enough (well more accurately stated, the problem is small/simple enough) that it doesn't matter about this lack of design skill in LLMs or agents. But don't confuse that for being able to design software in the more general use case.
Exactly, for the thing that has been done in Github 10000x times over, LLMs are pretty awesome and they speed up your job significantly (it's arguable if you would be better off using some abstraction already built if that's the case).
But try to do something novel and... they become nearly useless. Not like anything particularly difficult, just something that's so niche it's never been done before. It will most likely hallucinate some methods and call it a day.
As a personal anecdote, I was doing some LTSpice simulations and tried to get Claude Sonnet to write a plot expression to convert reactance to apparent capacitance in an AC sweep. It hallucinated pretty much the entire thing, and got the equation wrong (assumed the source was unit intensity, while LTSpice models AC circuits with unit voltage. This surely is on the internet, but apparently has never been written alongside the need to convert an impedance to capacitance!).
> Coding AIs design software better than me, review code better than me, find hard-to-find bugs better than me, plan long-running projects better than me, make decisions based on research, literature, and also the state of our projects better than me.
They don't do any of that better than me; they do it poorer and faster, but well enough for most of the time.
There will be a need. Don't worry. Most people still haven't figured out how to properly read and interpret instructions. So they build things incorrectly - with or without AI
Seriously. The bar is that low. When people say "AI slop" I just chuckle because it's not "AI" it's everyone. That's the general state of the industry.
So all you have to do is stay engaged, ask questions, and understand the requirements. Know what it is you're building and you'll be fine.
More than any other effect they have LLMs breed something called "learned helplessness". You just listed a few things it may stay better than you at, and a few things that it is not better than you at and never will be.
Planning long running projects and deciding are things only you can do well!! Humans manage costs. We look out for our future. We worry. We have excitement, and pride. It wants you to think none of these things matter of course, because it doesn't have them. It says plausible things at random, basically. It can't love, it can't care, it won't persist.
WHATEVER you do don't let it make you forget that it's a bag of words and you are someing almost infinitely more capable, not in spite of human "flaws" like caring, but because of them :)
Plus I think I've almost never see so little competition for what I think are the real prizes! Everyone's off making copies of copies of copies of the same crappy infrastructure we already have. They're busy building small inconsequential side projects so they can say they built something using an LLM.
> They're busy building small inconsequential side projects
Unironically, sending a program to build those for me have send me almost endless amount of time. I'm a pretty distracted individual, and pretty anal about my workflow/environment, so lots of times I've spent hours going into rabbit-holes to make something better, when I could have just sucked it up and do it the manual way instead, even if it takes mental energy.
Now, I can still do those things, but not spend hours, just a couple of minutes, and come back after 20-30 minutes to something that lets me avoid that stuff wholesale. Once you start stacking these things, it tends to save a lot of time and more importantly, mental energy.
So the programs by themselves are basically "small inconsequential side projects" because they're not "production worthy and web scale SaaS ready to earn money", but they help me and others who are building those things in a big way.
But isn't that exactly the kind of learned helplessness being discussed? As a fellow distracted individual, I have seen instant gratification erode all of my most prized hobbies and skills. Why read a book when I can scroll on my phone? My distress tolerance is lower than ever. LLMs feel like a bridge too far, for me anyway.
Nothing has been eroded for me, in fact it had the opposite effect. It's easier to get into new hobbies, easier to develop skills, I value reading on my own more than I did before. At least for me, LLMs act as multipliers of what I can and want to do, it hasn't removed my passion for music production, 3D, animation or programming one bit, if anything it's fueled those passions and let me do stuff within them faster and better.
Nothing I could make would be very good. So the only reason I would, say, write, is in order to write, not to have produced an essay. Hobbies are ways to pass time productively. If it took less time, it wouldn't be a better use of time, but a worse one.
It's not about being able to do more faster, but be able to faster get help doing what you wanted to do. For example, before LLMs, if I wanted to figure out how to do something with a specific analog synth I basically spent time reading manuals and browsing internet forums, piecing together whatever I could find into something actionable, sometimes slightly wrong, but at least in the right direction.
Nowadays, I fire off the LLM to figure it out for me, then try out what I get back, and I can move on to actually having fun playing on the synth, rather than trying to figure out how to do what I wanted to do.
The end goal for me with my hobbies is more or less the same, have fun. But for me the fun is not digging through manuals, it is to "do" or "use" or "perform" or whatever. I like music production because I like to make music, not because I like digging through manuals for some arcane knowledge.
The point is "things that used to take me hours, can now be done by a magic computer program in the background, while I do other things". It's applicable for small unix utilities I create to make my development UX better, it's applicable for when I'm doing music production and it's applicable in a wide-range of tasks both professionally and for my hobbies.
It saves me from stuff I find boring yet necessary, so I can focus more on the fun stuff. I guess this was the overall point I was trying to make in this comment-chain.
Yea I've been seeing very similar behavior from people. They think of themselves as static, unchanging, uncreative but view LLMs as some kind of unrelenting and inevitable innovative force...
I think it's people's anxieties and fears about the uncertainty about the value of their own cognitive labor demoralizing them and making them doubt their own self-efficacy. Which I think is an understandable reaction in the face of trillion dollar companies frothing at the mouth to replace you with pale imitations.
You are still in denial of what an LLM actually is capable of in the near-mid term.
In the current architecture there are mathmatical limitations on what it can do with information. However, tool use and external orchestration allow it to work around many (maybe all) those limitations.
The current models have brittle parts and some bad tendencies.. but they will continue to eat up the executive thought ladder.
I think it is better to understand this and position yourself higher and higher on that chain while learning what are the weak areas in the current generation of models.
Your line of thinking is like hiding in a corner while the water is rising. You are right, it is a safe corner, but probably not for long.
I don't think the limitations on what it can do are mathematical at all. It has no, faith, no conviction, no sense of self. No philosophy, no ability to learn. How could it undertake a major effort?
Just so we are clear, you are saying you don't use it at all, but you are providing advice about it? Specifically detailing with certainty that the current state of the art has or doesn't have certain traits or abilities.
Yes. I'm not providing advice on how to use it, I'm providing advice on whether or not to use it. A million people cried out that I would be obsolete. I would be replaced: left behind. Career suicide one said LOL.
I think I'm the perfect person to be qualified to stand up and say "if they tell you you can't live without it, they are lying to your face." Only someone who has lived without it as I have would be in a position to know
Where the hell was all this fear when the push for open source everything got fully underway? When entire websites were being spawned and scaffolded with just a couple lines of code? Do we not remember all those impressive tech demos of developers doing massive complex thing with "just one line of code"? How did we not just write software for every kind of software problem that could exist by now?
How has free code, developed by humans, become more available than ever and yet somehow we have had to employ more and more developers? Why didn't we trend toward less developers?
It just doesn't make sense. AI is nothing but a snippet generator, a static analyzer, a linter, a compiler, an LSP, a google search, a copy paste from stackoverflow, all technologies we've had for a long time, all things developers used to have to go without at some point in history.
> Coding AIs design software better than me, review code better than me, find hard-to-find bugs better than me, plan long-running projects better than me, make decisions based on research, literature, and also the state of our projects better than me
Perfect economic substitution in coding doesn't happen for a long time. Meanwhile, AI appears as an amplifier to the human and vice versa. That the work will change is scary, but the change also opens up possibilities, many of them now hard to imagine.
Stop freaking out. Seriously. You're afraid of something completely ridiculous.
It is certainly more eloquent than you regarding software architecture (which was a scam all along, but conversation for another time).
It will find SOME bugs better than you, that's a given.
Review code better than you? Seriously? What you're using and what you consider code review?
Assume I could identify one change broke production and you reviewed the latest commit. I am pinging you and you better answer. Ok, Claude broke production, now what?
Can you begin to understand the difference between you and the generative technology?
When you hop on the call, you will explain to me with a great deal of details what you know about the system you built, and explain decision making and changes over time. You'll tell about what worked and what didn't. You will tell about the risks, behavior and expectations. About where the code runs, it's dependencies, users, usage patterns, load, CPU usage and memory footprint, you could probably tell what's happening without looking at logs but at metrics.
With Claude I get: you're absolutely right! You asked about what it WAS, but I told you about what it WASN'T! MY BAD.
Knowledge requires a soul to experience and this is why you're paid.
We use code rabbit and it's better than practically any human I've worked with at a number of code review tasks, such as finding vulnerabilities, highlighting configuration issues, bad practices, etc. It's not the greatest at "does this make sense here" type questions, but I'd be the one answering those questions anyway.
Yeah, maybe the people I've worked with suck at code reviews, but that's pretty normal.
Not to say your answer is wrong. I think the gist is accurate. But I think tooling will get better at answering exactly the kind of questions you bring up.
Also, someone has to be responsible. I don't think the industry can continue with this BS "AI broke it." Our jobs might devolve into something more akin to a SDET role and writing the "last mile" of novel code the AI can't produce accurately.
> We use code rabbit and it's better than practically any human
code rabbit does find things occasionally, but it also calls things 'critical' that arent and flags issues that dont actually exist and even lies in replies sometimes...
it also is extremely verbose to the point of being slog to go through... and the haikus: they are so cringe and infantilizing...
maybe its our config, but code rabbit has been underwhelming...
Yes, seriously (not OP). Sometimes it's dumb as rocks, sometimes it's frighteningly astute.
I'm not sure at which point of the technology sigmoid curve we find ourselves (2007 iPhone or 2017 iPhone?) but you're doing yourself a disservice to be so dismissive
Copilot reviews are enabled company wide and comments must be resolved manually. I wish I could be so dismissive lol
I cannot, literally do not have the ability to be dismissive
The AI is pretty scary if you think most of software engineering is about authoring individual methods and rubber ducking about colors of paint and brands of tools.
Once you learn that it's mostly about interacting with a customer (sometimes this is yourself), you will realize the AI is pretty awful at handling even the most basic tasks.
Following a product vision, selecting an appropriate architecture and eschewing 3rd party slop are examples of critical areas where these models are either fundamentally incapable or adversely aligned. I find I have to probe ChatGPT very hard to get it to offer a direct implementation of something like a SAML service provider. This isn't a particularly difficult thing to do in a language like C# with all of the built in XML libraries, but the LLM will constantly try to push you to use 3rd party and cloud shit throughout. If you don't have strong internal convictions (vision) about what you really want, it's going to take you for a ride.
One other thing to remember is that our economies are incredibly efficient. The statistical mean of all information in sight of the LLMs likely does not represent much of an arbitrage opportunity at scale. Everyone else has access to the same information. This also means that composing these systems in recursive or agentic styles means you aren't gaining anything. You cannot increase the information content of a system by simply creating another instance of the same system and having it argue with itself. There usually exists some simple prompt that makes a multi agent Rube Goldberg contraption look silly.
> I’m basically just the conductor of all those processes.
"Basically" and "just" are doing some heroic weight lifting here. Effectively conducting all of the things an LLM is good at still requires a lot of experience. Making the constraints live together in one happy place is the hard part. This is why some of us call it "engineering".
This reads like shilling/advertisement.. Coding AIs are struggling for anything remotely complex, make up crap and present it as research, write tests that are just "return true", and won't ever question a decision you make.
Those twenty engineers must not have produced much.
I think part of what is happening here is that different developers on HN have very different jobs and skill levels. If you are just writing a large volume of code over and over again to do the same sort of things, then LLMs probably could take your job. A lot of people have joined the industry over time, and it seems like the intelligence bar moved lower and lower over time, particularly for people churning out large volumes of boilerplate code. If you are doing relatively novel stuff, at least in the sense that your abstractions are novel and the shape of the abstraction set is different from the standard things that exist in tutorials etc online, then the LLM will probably not work well with your style.
So some people are panicking and they are probably right, and some other people are rolling their eyes and they are probably right too. I think the real risk is that dumping out loads of boilerplate becomes so cheap and reliable that people who can actually fluently design coherent abstractions are no longer as needed. I am skeptical this will happen though, as there doesn’t seem to be a way around the problem of the giant indigestible hairball (I.e as you have more and more boilerplate it becomes harder to remain coherent).
Indeed, discussions on LLMs for coding sound like what you would expect if you asked a room full of people to snatch up a 20 kg dumbbell once and then tell you if it's heavy.
> I think the real risk is that dumping out loads of boilerplate becomes so cheap and reliable that people who can actually fluently design coherent abstractions are no longer as needed.
Cough front-end cough web cough development. Admittedly, original patterns can still be invented, but many (most?) of us don't need that level of creativity in our projects.
Absolutely this, and TFA touches on the point about natural language being insufficiently precise:
AI can write you an entire CRUD app in minutes, and with some back-and-forth you can have an actually-good CRUD app in a few hours.
But AI is not very good (anecdotally, based on my experience) at writing fintech-type code. It's also not very good at writing intricate security stuff like heap overflows. I've never tried, but would certainly never trust it to write cryptography correctly, based on my experience with the latter two topics.
All of the above is "coding", but AI is only good at a subset of it.
Generating CRUD is like solving cancer in mice, we already have a dizzying array of effective solutions… Ruby on Rails, Access 97, model first ORMs with GUI mappers. SharePoint lets anyone do all the things easily.
The issue is and always has been maintenance and evolution. Early missteps cause limitations, customer volume creates momentum, and suddenly real engineering is needed.
I’d be a lot more worried about our jobs if these systems were explaining to people how to solve all their problems with a little Emacs scripting. As is they’re like hyper aggressive tech sales people, happy just to see entanglements, not thinking about the whole business cycle.
Go with Laravel and some admin packages and you generate CRUD pages in minutes. And I think with Django, that is builtin.
But I don’t think I’ve seen pure CRUD on anything other than prototype. Add an Identity and Access Management subsystem and the complexity of requirements will explode. Then you add integration to external services and legacy systems, and that’s where the bulk of the work is. And there’s the scalability issue that is always looming.
Creating CRUD app is barely a level over starting a new project with the IDE wizard.
>Creating CRUD app is barely a level over starting a new project with the IDE wizard.
For you, maybe. But for a non-progrmamer who's starting a business or just needs a website it's the difference between hiring some web dev firm and doing it themselves.
> it's the difference between hiring some web dev firm and doing it themselves.
anecdote but i've had a lot of acquaintances who started at both "hiring some web dev firm" and "doing it themselves" with results largely being the same: "help me fix this unmaintainable mess and i will pay you x"...
jmo but i suspect llms will allow for the later to go further before the "help me" phase but i feel like that aint going away completely...
Just like my previous comments, much depends on the specifics.
My wife's sister and her husband run a small retail shop in $large_city. My sister-in-law taught herself how to set up and modify a website with a shopify storefront largely with LLM help. Now they take online orders. I've looked at the code she wrote and it's not pretty but it generally works. There will probably never be a "help me fix this unmaintainable mess and I will pay you" moment in the life of that business.
The crux of my point is this: In 2015 she would have had to hire somebody to do that work.
This segment of the software industry is where the "LLMs will take our jerbs" argument is coming from.
The people who say "AI is junk and it can't do anything right" are simply operating in a different part of the industry.
> different developers on HN have very different jobs and skill levels.
Definitely this. When I use AIs for web development they do an ok job most of the time. Definitely on par with a junior dev.
For anything outside of that they're still pretty bad. Not useless by any stretch, but it's still a fantasy to think you could replace even a good junior dev with AI in most domains.
I am slightly worried for my job... but only because AI will keep improving and there is a chance it will be as good as me one day. Today it's not a threat at all.
Yea, LLMs produce results on par with what I would expect out of a solid junior developer. They take direction, their models act as the “do the research” part, and they output lots of code: code that has to be carefully scrutinized and refined. They are like very ambitious interns who never get tired and want to please, but often just produce crap that has to be totally redone or refactored heavily in order to go into production.
If you think LLMs are “better programmers than you,” well, I have some disappointing news for you that might take you a while to accept.
> LLMs produce results on par with what I would expect out of a solid junior developer
This is a common take but it hasn't been my experience. LLMs produce results that vary from expert all the way to slightly better than markov chains. The average result might be equal to a junior developer, and the worst case doesn't happen that often, but the fact that it happens from time to time makes it completely unreliable for a lot of tasks.
Junior developers are much more consistent. Sure, you will find the occasional developer that would delete the test file rather than fixing the tests, but either they will learn their lesson after seeing your wth face or you can fire them. Can't do that with llms.
I think any further discussion about quality just needs to have the following metadata:
- Language
- Total LOC
- Subject matter expertise required
- Total dependency chain
- Subjective score (audited randomly)
And we can start doing some analysis. Otherwise we're pissing into ten kinds of winds.
My own subjective experience is earth shattering at webapps in html and css (because I'm terrible and slow at it), and annoyingly good but a bit wrong usually in planning and optimization in rust and horribly lost at systems design or debugging a reasonably large rust system.
I agree in that these discussions (this whole hn thread tbh) are seriously lacking in concrete examples to be more than holy wars 3.0.
Besides one point: junior developers can learn from their egregious mistakes, llms can't no matter how strongly worded you are in their system prompt.
In a functional work environment, you will build trust with your coworkers little by little. The pale equivalent in LLMs is improving system prompts and writing more and more ai directives that might or might not be followed.
This seems to be one of the huge weaknesses of current LLMs: Despite the words "intelligence" and "machine learning" we throw around, they aren't really able to learn and improve their skills without someone changing the model. So, they repeat the same mistakes and invent new mistakes by random chance.
If I was tutoring a junior developer, and he accidentally deleted the whole source tree or something egregious, that would be a milestone learning point in his career, and he would never ever do it again. But if the LLM does it accidentally, it will be apologetic, but after the next context window clear, it has the same chances of doing it again.
> Besides one point: junior developers can learn from their egregious mistakes, llms can't no matter how strongly worded you are in their system prompt.
I think if you set off an LLM to do something, and it does a "egregious mistake" in the implementation, and then you adjust the system prompt to explicitly guard against that or go towards a different implementation and you restart from scratch again yet it does the exact same "egregious mistake", then you need to try a different model/tool than the one you've tried that with.
It's common with smaller models, or bigger models that are heavily quanitized that they aren't great at following system/developer prompts, but that really shouldn't happen with the available SOTA models, I haven't had something ignored like that in years by now.
And honestly this is precisely why I don't fear unemployment, but I do fear less employment overall. I can learn and get better and use LLMs as a tool. So there's still a "me" there steering. Eventually this might not be the case. But if automating things has taught me anything, it's that removing the person is usually such a long tail cost that it's cheaper to keep someone in the loop.
But is this like steel production or piloting (few highly trained experts are in the loop) or more like warehouse work (lots of automation removed any skills like driving or inventory work etc).
> If you are just writing a large volume of code over and over again
But why would you do that? Wouldn't you just have your own library of code eventually that you just sell and sell again with little tweaks? Same money for far less work.
People, at least novice developers, tend to prefer fast and quick boilerplate that makes them look effective, over spending one hour sitting just thinking and designing, then implementing some simple abstraction. This is true today, and been true for as long as I've been in programming.
Besides, not all programming work can be abstracted into a library and reused across projects, not because it's technically infeasible, but because the client doesn't want to, cannot for legal reasons or the developer process at the client's organization simply doesn't support that workflow. Those are just the reasons from the top of my head, that I've encountered before, and I'm sure there is more reasons.
But people don't stay novices after years/decades. Of course when you write the boilerplate for the 20x time maybe you still accept that, but when you write it for the 2000x time, I bet you do the lazy thing and just copy it.
> cannot for legal reasons or ...
Sure, you can't copy trade secrets, but that's also not the boilerplate part. Copying e.g. a class hierarchy and renaming all the names and replacing the class contents that represent the domain, won't be a legal problem, because this is not original in the first place.
> But people don't stay novices after years/decades
Some absolutely do. I know programmers who entered web development at the same time as me, and now after decades they're still creating typical CRUD applications for whatever their client today is, using the same frameworks and languages. If it works, makes enough money and you're happy, why change?
> Copying e.g. a class hierarchy and renaming all the names and replacing the class contents that represent the domain, won't be a legal problem, because this is not original in the first place.
Some code you produce for others definitively fall under their control, but obviously depends on the contracts and the laws of the country you're in. But I've written code for others that I couldn't just "abstract into a FOSS library and use in this project", even if it wasn't trade secrets or what not, just some utility for reducing boilerplate.
> "abstract into a FOSS library and use in this project"
That is not what I meant. My idea was more like "copy ten lines from this project, then lines from that project, the class from here, but replace every line before the commit ...".
I shouldn't have used the word library, as I did not mean output from the linker, but rather a colloquial meaning of a loose connection of snippets.
That’s a very good point I hadn’t heard explained that way before. Makes a lot of sense and explains a lot of the circular debates about AI that happen here daily.
> Existing abstractions are sufficient for building any software you want.
Software that doesn't need new abstractions is also already existing. Everything you would need already exists and can be bought much more cheaply than you could do it yourself. Accounting software exists, unreal engine exists and many games use it, why would you ever write something new?
>Software that doesn't need new abstractions is also already existing
This isn't true due to the exponential growth of how many ways you can compose existing abstractions. The chance that a specific permutation will have existing software is small.
I'm supposing that nobody who has a job is producing abstractions that are always novel, but there may be people who find abstractions that are novel for their particular field because it is something most people in that field are not familiar with, or that come up with novel abstractions (infrequently) that improve on existing ones.
The new abstraction is “this corporation owns this IP and has engineers who can fix and extend it at will”. You can’t git clone that.
But if there is something off the shelf that you can use for the task at hand? Great! The stakeholders want it to do these other 3000 things before next summer.
Software development is a bit like chess. 1. e4 is an abstraction available to all projects, 3. Nc3 is available to 20% of projects, while 15. Nxg5 is unique to your own project.
Or, abstractions in your project form a dependency tree, and the nodes near the root are universal, e.g. C, Postgres, json, while the leaf nodes are abstractions peculiar to just your own project.
The possible chess moves is already known ahead of time. Just because an AI can't make up a move like Np5 as a human could do, that doesn't mean anything AI can't play chess. It will be fine with just using the existing moves that have been found so far. The idea that we still need humans to come up with new chess moves is not a requirement for playing chess.
No it doesn’t read like shilling and advertisement, it’s tiring hearing people continually dismiss coding agents as if they have not massively improved and are driving real value despite limitations and they are only just getting started. I’ve done things with Claude I never thought possible for myself to do, and I’ve done things where Claude made the whole effort take twice as long and 3x more of my time. It’s not like people are ignoring the limitations, it’s that people can see how powerful the already are and how much more headroom there is even with existing paradigms not to mention the compute scaling happening in 26-27 and the idea pipeline from the massive hoarding of talent.
When prices go down or product velocity goes up we'll start believing in the new 20x developer. Until then, it doesn't align with most experiences and just reads like fiction.
You'll notice no one ever seems to talk about the products they're making 20x faster or cheaper.
AI boosters? Like people are planted by Sam Altman like the way they hire crowds for political events or something? Hey! Maybe I’m AI! You’re absolutely right!
In seriousness: I’m sure there are projects that are heavily powered by Claude, myself and a lot of other people I know use Claude almost exclusively to write and then leverage it as a tool when reviewing. Almost everyone I hear that has this super negative hostile attitude references some “promise” that has gone unfulfilled but it’s so silly: judge the product they are producing and maybe just maybe consider the rate of progress to _guess_ where things are heading
I never said "planted", that is your own assumption, albeit a wrong one. I do respect it though, as it is at least a product of a human mind. But you don't have to be "planted" to champion an idea, you are clearly championing it out of some kind of conviction, many seem to do. I was just giving you a bit of reality check.
If you want to show me how to "guess where things are heading" / I am actually one of the early adopters of LLMs and have been engineering software professionally for almost half my life now. Why do you think I was an early adopter?
Because I was skeptical or afraid of that tech? No, I was genuinely excited. Yes you can produce mountains of code, even more so if you were already an experienced engineer, like myself for example.
Yes you can even get it to produce somewhat acceptable outputs, with a lot of effort at prompting it and fatigue that comes with it. But at the end of the day, as an experienced engineer, I am not being more productive with it, I will end up being less productive because of all the sharp edges I have to take care of, all the sloppily produced code, unnecessary bloat, hallucinated or injected libraries etc.
Maybe for folks who were not good at maths or had trouble understanding how computers work this looks like a brave new world of opportunities. Surely that app looks good to you, how bad can it be? Just so you and other such vibe-coders understand, here is a parallel.
It is actually fairly simple for a group of aviation enthusiasts to build a flying airplane. We just need to work out some basic mechanics, controls and attach engines. It can be done, I've seen a couple of documentaries too. However, those planes are shit. Why? Because me and my team of enthusiast dont have the depth of knowledge of a team of aviation engineers to inform my decisions.
What is the tolerance for certain types of movements, what kind of materials do I need to pick, what should be my maintenance windows for various parts etc. There are things experts can decide on almost intuitively, yet with great precision, based on their many years of craft and that wonderful thing called human intelligence. So my team of enthusiasts puts together an airplane. Yeah it flies. It can even be steered. It rolls, pitches and yawns. It takes off and lands. But to me it's a black-box, because I don't understand many, many factors, forces, pressures, tensors, effects etc that are affecting an airplane during it's flight and takeoff. I am probably not even aware WHAT I should be aware of. Because I dont have that deep educaiton about mechanical engineering, materials, aerodynamics etc. Neither does my team. So my plane, while impressive to me and my team, will never take off commercially, not unless a team of professionals take it over and remakes it to professional standards. It will probably never even fly in a show. And if me or someone on my team dies flying it, you guessed it - our insurance sure as hell won't cover the costs.
So what you are doing with Claude and other tools, while it may look amazing to you, is not that impressive to the rest of us, because we can see those wheels beginning to fall off even before your first take off. Of course, before I can even tell that, I'd have to actually see your airplane, it's design plans etc. So perhaps first show us some of those "projects heavily powered by Claude" and their great success, especially commercial one (otherwise its a toy project), before you talk about them.
The fact that you are clearly not an expert on the topic of software engineering should guide you here - unless you know what you are talking about, it's better to not say anything at all.
How would you know whether he is an expert on the topic of software engineering or not?
For all I know, he is more competent than you; he figured out how to utilize Claude Code in a productive way, which is a point for him.
I'd have to guess whether you are an expert working on software not well suited for AI, or just average with a stubborn attitude towards AI and potentially not having tried the latest generation of models and agentic harnesses.
I think it's worth framing things back to what we're reacting to. The top poster said:
> I really really want this to be true. I want to be relevant. I don’t know what to do if all those predictions are true and there is no need (or very little need) for programmers anymore.
The rest of the post is basically their human declaration of obsolescence to the programming field. To which someone reacted by saying that this sounds like shilling. And indeed it does for many professional developers, including those that supplement their craft with LLMs. Declaring that you feel inadequate because of LLMs only reveals something about you. Defending this position is a tell that puts anyone sharing that perspective in the same boat: you didn't know what you were doing in the first place. It's like when someone who couldn't solve the "invert a binary tree" problem gets offended because they believed they were tricked into an impossible task. No, you may be a smart person that understands enough of the rudiment of programming to hack some interesting scripts, but that's actually a pretty easy problem and failing to solve it indeed signals that you lack some fundamentals.
> Considering those views are shared by a number of high profile, skilled engineers, this is obviously no basis for doubting someone's expertise.
I've read Antirez, Simon Willison, Bryan Cantrill, and Armin Ronacher on how they work or want to work with AI. From none I've got this attitude that they're no longer needed as part of the process.
> Considering those views are shared by a number of high profile, skilled engineers, this is obviously no basis for doubting someone's expertise
Again, a lot of fluff, a lot of of "a number ofs", "highly this, highly that". But very little concrete information. What happened to the pocket PhDs promised for this past summer? Where are the single-dude billion dollar companies built with AI tools ? Or even a multiple-dudes billion dollar companies ? What are you talking about?
I've yet to see it from someone who isn't directly or indirectly affiliated with an organisation that would benefit from increased AI tool adoption. Not saying it's impossible, but...
Whereas there are what feels like endless examples of high profile, skilled engineers who are calling BS on the whole thing.
You can say the same about people saying the opposite. I haven’t heard from a single person who says AI can’t write code that does not a financially interest directly or indirectly in humans writing code.
That seems rather disingenuous to me. I see many posts which clearly come from developers like you and me who are happy with the results they are getting.
Every time people on here comment something about "shilling" or "boosters". It would seem to me that in the rarest of cases someone shares their opinion to profit from it, while you act like that is super common.
Right: they disagree with me and so must not know what they’re talking about. Hey guess how I know neither of you are all as good as you think you are: your egos! You know what the brightest people at the top of their respective fields have in common? They tend not to think that new technologies they don’t understand how to use are dumb and they don’t think everyone who disagrees with them is dumb!
> you are clearly not an expert on the topic of software engineering should guide you here - unless you know what you are talking about, it's better to not say anything at all.
Yikes, pretty condescending. Also wrong!
IMO you are strawmanning pretty heavily here.
Believe it or not, using Claude to improve your productivity is pretty dissimilar to vibe coding a commercial airplane(?) which I would agree is probably not FAA approved.
I prefer not to toot my own horn, but to address an idea you seem to have that I don’t know math or CS(?) I have a PhD in astrophysics and a decade of industry experience in tech and other domains so I’m fairly certain I know how math and computers work but maybe not!
I’m an expert in what I do. A professional, and few people can do what I do. I have to say you are wrong. AI is changing the game. What you’ve written here might’ve been more relevant about 9 months ago, but everything has changed.
Right I’m a bot made to promote AI like half the people on this thread.
I don’t know if you noticed a difference from other hype cycles but other ones were speculative. This one is also speculative but the greater divide is that the literal on the ground usefulness of AI is ALREADY going to change the world.
The speculation is that the AI will get better and will no longer need hand holding.
I'm having a lot of trouble understanding what you're trying to convey. You say there's a difference from previous "speculation" but also that it's still speculation. Then you go on to write "ALREADY going to" which is future tense (speculation), even clarifying what the speculation is.
So let me explain it more clearly. AI as it is now is already changing the game. It will reduce the demand of swes across every company as an eventuality if we hold technological progress fixed. There is no speculation here. This comes from on the ground evidence from what I see day to day and what I do and my experience pair programming things from scratch with AI.
The speculation is this: if we follow the trendlines of AI improvement for the past decade and a half, the projection of past improvement indicates AI will only get better and better. It’s a reasonable speculation, but it is nonetheless speculative. I wouldn’t bet my life on continuous improvement of AI to the point of AGI but it’s now more than ever before a speculation that is not unrealistic.
Nice slop response. This is the same thing said about blockchain and NFTs, same schtick, different tech. The only thing "AI" has done is convince some people that it's a magical being that knows everything. Your comments seem to be somewhere on that spectrum. And, sure what if it isn't changing the world for the better, and actually makes things much worse? You're probably okay with that too, I guess, as long as your precious "AI" is doing the changing.
We've seen what social media and every-waking-hour access to tablets and the internet has done to kids - so much harm that some countries have banned social media for people under a certain age. I can see a future where "AI" will also be banned for minors to use, probably pretty soon too. The harms from "AI" being able to placate instead of create should be obvious, and children shouldn't be able to use it without adult supervision.
>The speculation is that the AI will get better and will no longer need hand holding.
This is nonsense. No AI is going to produce what someone wants without telling it exactly what to do and how to do it, so yes, it will always need hand holding, unless you like slurping up slop. I don't know you, if you aren't a bot, you might just be satisfied with slop? It's a race to the bottom, and it's not going to end up the way you think it will.
>This is nonsense. No AI is going to produce what someone wants without telling it exactly what to do and how to do it, so yes, it will always need hand holding, unless you like slurping up slop. I don't know you, if you aren't a bot, you might just be satisfied with slop? It's a race to the bottom, and it's not going to end up the way you think it will.
You're not thinking clearly. A couple years ago we didn't even have AI who could do this, then chatGPT came out we had AI who could barely do it, then we had AI who could do simple tasks with A lot of hand holding, now we have AI who can do complex human tasks with minimal hand holding. Where do you think the trendline is pointing.
Your hypothesis is going against all evidence. It's more wishful thinking and irrational. It's a race to the bottom because you wish it will be a race to the bottom, and we both know the trendline is pointing in the opposite direction.
>We've seen what social media and every-waking-hour access to tablets and the internet has done to kids - so much harm that some countries have banned social media for people under a certain age. I can see a future where "AI" will also be banned for minors to use, probably pretty soon too. The harms from "AI" being able to placate instead of create should be obvious, and children shouldn't be able to use it without adult supervision.
I agree AI is bad for us. My claim is it's going to change the world and it is already replacing human tasks. That's all. Whether that's good or bad for us is an ORTHOGANOL argument.
I use AI every day, and it's honestly crap. No it isn't significantly improving, it's hitting a wall. Every new model release is getting less and less good, so no, the "trendline" is not going up as much as you seem to think it is. It's plateaued. The only way "AI" is going to change the world is if stupid people put it in places that it really shouldn't be, thinking it will solve problems and not create even more problems.
Proof of what? Should you also have to prove you are not a bot sponsored by short-sellers? It’s all so so silly, anti-AI crowds on HN rehash so many of the same tired arguments it’s ridiculous:
- bad for environment: how? Why?
- takes all creative output and doesn’t credit: common crawl has been around for decades and models have been training for decades, the difference is that now they’re good. Regurgitating training data is a known issue for which there are mitigations but welcome to the world of things not being as idealistic as some Stallman-esque hellscape everyone seems to want to live in
- it’s bad and so no one should use it and any professionals who do don’t know what they’re doing: I have been so fortunate to personally know some of the brightest minds on this planet (Astro departmentments, AI research labs) and majority of them use AI for their jobs.
>Should you also have to prove you are not a bot sponsored by short-sellers?
On a 35 day-old account, yes. Anything "post-AI" is suspect now.
The rest of your comment reads like manufactured AI slop, replying to things I never even wrote in my one sentence comment. And no surprise coming from an account created 1 day ago.
I think it’s quite obvious I’m not writing AI slop.
The latest chatgpt for example will produce comments that are now distinguishable from the real thing only because they’re much better written. It’s insane that the main visible marker rn is that the arguments and writings it crafts are superior then what your average joe can write.
My shit writing can’t hold a candle and that’s pretty obvious. AI slop is not accepted here but I can post an example of what AI slop will now look like, if AI responded to you it would look like this:
Fair to be skeptical of new accounts. But account age and “sounds like AI” are not workable filters for truth. Humans can write like bots, bots can write like humans, and both can be new. That standard selects for tenure, not correctness.
More importantly, you did not engage any claim. If the position is simply “post-AI content from new accounts is suspect,” say that as a moderation concern. But as an argument, suspicion alone does not refute anything.
Pick one concrete claim and say why it is wrong or what evidence would change your mind. Otherwise “this reads like slop” is just pattern matching. That is exactly the failure mode being complained about.
I accused another user of writing AI slop in this specific thread, and here you are inserting yourself as if you are replying to comment I made to the other user. You certainly seem desperate to boost "AI" as much as you can. Your 37 day old account is also just as suspect as their 3 day old account. I'm not engaging with you any more so replying is kind of pointless.
Obviously not troll, I know I’m bragging. But I have to emphasize that it is not some stupid oh “only domain experts know AI is shit. Everyone else is too stupid to understand how bad it is” That is patently wrong.
Few people can do what I do and as a result I likely make more money than you. But now with AI… everyone can do what I do. It has leveled the playing field… what I was before now matters fuck all. Understand?
I still make money right now. But that’s unlikely to last very long. I fully expect it to disappear within the next decade.
You are wrong. People like yourself will likely be smart enough to stay well employed into the future. It's the folks who are arguing with you trying to say that AI is useless who will quickly lose their jobs. And they'll be all shocked Pikachu face when they get a pink slip while their role gets reassigned to an AI agent
> It's the folks who are arguing with you trying to say that AI is useless who will quickly lose their jobs.
Why is it that in every hype there are always the guys like you that want to punish the non-believers? It's not enough to be potentially proven correct, your anger requires the demise of the heretics. It was the same story for cryptocurrencies.
He/she is probably one of those poor souls working for an AI-wrapper-startup who received a ton of compensation in "equity", which will be worth nothing when their founders get acquihired, Windsurf style ;) But until then, they get to threaten us all with the impending doom, because hey, they are looking into the eye of the storm, writing Very Complex Queries against the AI API or whatever...
Isn’t this the same type of emotional response he’s getting accused for? You’re speculating that he will be “punished” just as he speculated for you.
There’s emotions on both sides and the goal is to call it out, throw it to the side and cut through into the substance. The attitude should be: Which one of us is actually right? Rather than: I’m right and you’re a fucking idiot attitude I see everywhere.
Mate, I could not care less if he/her got "punished" or not. I was just assuming what might be driving someone to go and try and answer each and every one of my posts with very low quality comments, reeking of desperation and "elon-style" humour (cheap, cringe puns). You are assuming too much here.
Not too dissimilar to you. I wrote long rebuttals to you points and you just descended into put downs, stalking and false accusations. You essentially told me to fuck off from all of HN in one of your posts.
Bro idk why you waste your time writing all this. No one cares that you were an early adopter, all that means is that you used the rudimentary LLM implementations that were available from 2022-2024 which are now completely obselete. Whatever experience you think you have with AI tools is useless because you clearly haven't kept up with the times. AI platforms and tools have been changing quickly. Every six months the capabilities have massively improved.
Next time before you waste ten minutes typing out these self aggrandizing tirades maybe try asking the AI to just write it for you instead
Maybe he's already ahead of you by not using current models, 2026 models are going to make 2025 models completely obsolete, wasting time on them is dumb.
This is such a fantastic response. And outsiders should very well be made aware what kind of plane they are stepping into. No offence to the aviation enthusiasts in your example but I will do everything in my power to avoid getting on their plane, in the same way I will do everything in my power to avoid using AI coded software that does anything important or critical...
> but I will do everything in my power to avoid getting on their plane
speaking of airplanes... considering how much llm usage is being pushed top-down in many places, i wonder how long until some news drops of some catastrophic one-liner got through via llm generated code...
"Littered" is a great verb to use here. Also I did not ask for a deviated proxy non-measure, like how many people who are choking themselves to death in a meaningless bullshit job are now surviving by having LLMs generate their spreadsheets and presentations. I asked for solid proof of succesful, commercial products built up by dreaming them up through LLMs.
The proof is all around you. I am talking about software professionals not some bullshit spread sheet thing.
What I’m saying is this: From my pov Everyone is using LLMs to write code now. The overwhelming majority of software products in existence today are now being changed with LLM code.
The majority of software products being created from scratch are also mostly LLM code.
This is obvious to me. It’s not speculation, where I live and where I’m from and where I work it’s the obvious status quo. When I see someone like you I’m thinking because the change happened so fast you’re one of the people living in a bubble. Your company and the people around you haven’t started using it because the culture hasn’t caught up.
Wait until you have that one coworker who’s going at 10x speed as everyone else and you find out it’s because of AI. That is what will slowly happen to these bubbles. To keep pace you will have to switch to AI to see the difference.
I also don’t know how to offer you proof. Do you use google? If so you’ve used products that have been changed by LLM code. Is that proof? Do you use any products built by a start up in the last year? The majority of that code will be written by an LLM.
> Your company and the people around you haven’t started using it because the culture hasn’t caught up.
We have been using LLMs since 2021, if I havent repeated that enough in these threads. What culture do I have to catch up with? I have been paying top tier LLM models for my entire team since it became an option. Do you think you are proselytizing to the un-initiated here? That is a naive view at best. My issue is that the tools are at best a worse replacement for the pre-2019 google search and at worst a huge danger in the hands of people who dont know what they are doing.
Doesn’t make sense to me. If it’s bad why pay for the tool?
Obviously your team disagrees that it’s a worse replacement for google or else why demand it against your will?
> at worst a huge danger in the hands of people who dont know what they are doing.
I agree with this. But the upside negates this and I agree with your own team on that.
Btw if you’re paying top dollar for AI.. your developers are unlikely using it as a google search replacement. At top dollar AI is used as an agent. What it ends up doing is extremely different from a google search in this mode. That may be good or bad but it is a distinctly different outcome then a google search and that makes your google analogy ill fitted to what your team is actually using it for.
Have you had your head in the sand for the past two years?
At the recent AWS conference, they were showcasing Kiro extensively with real life products that have been built with it. And the Amazon developers all allege that they've all been using Kiro and other AI tools and agents heavily for the past year+ now to build AWS's own services. Google and Microsoft have also reported similar internal efforts.
The platforms you interact with on a daily basis are now all being built with the help of AI tools and agents
If you think no one is building real commercial products with AI then you are either blind or an idiot or both. Why don't you just spend two seconds emailing your company AWS ProServe folks and ask them, I'm sure they'll give you a laundry list of things they're using AI for internally and sign you up for a Kiro demo as well
Amazon, Google and Microsoft are balls deep invested in AI, a rational person should draw 0 conclusions in them showcasing how productive they are with it.
I'd say it's more about the fear of their $50billion+ investments not paying off is creeping up on them.
It’s ok to have this prior but these are not speculative tools and capabilities, they exist today. If you remain unimpressed by them that’s fine, but to deny real people (not bots!) and real companies (we measure lots of stuff, I’ve seen the data at a large MAANG and have used their internal and external tools) get serious benefits _today_ and we still have about 4 more orders of magnitude to scale _existing_ paradigms, the writing on the wall is so obvious. It’s fine and reasonable to be skeptical and there are so many serious serious societal risks and issues to worry about and champion but to me if your position is akin to “this is all hype” it makes absolutely no sense to me
I'm sure you're interacting with a ton of tools built via agents, ironically even in software engineering people are trying to human-wash AI code due to anti-AI bias by people who should know better (if you think 100% of LLM outputs are "slop" with no quality consideration factored in, you're hopelessly biased). The commercialized seems like an arbitrary and pointless bar, I've seen some hot garbage that's "commercialized" and some great code that's not.
> I'm sure you're interacting with a ton of tools built via agents, ironically even in software engineering people are trying to human-wash AI code due to anti-AI bias
Please just for fun - reach out to for example Klarna support via their website and tell me how much of your experience can be attributed to an anti-AI bias and how much to the fact that the LLMs are a complete shit for any important production use cases.
My man here is reaching out to Klarna Support, this tells a LOT about his life decision making skills which clearly shine through as well in his comments on the topic of AI
You’ve never read Simon Willison’s blog? His repo is full of work that he’s created with LLM’s. He makes money off of them. There are plenty of examples you just need to look.
The paradigm shift hit the world like a wall. I know entire teams where the manager thinks AI is bullshit and the entire team is not allowed to use AI.
I love coding. But reality is reality and these fools just aren’t keeping pace with how fast the world is changing.
Or we're in another hype cycle and billions of dollars are being pumped in to sustain the current bubble with a lot of promises about how fast the world is changing. Doesn't mean AI can't be a useful tool.
When people say “hype cycle” that can mean so many different things. That valuations are too high and many industry “promises” are wrong is maybe true but to me it’s irrelevant, this isn’t speculative, I think most posters who are positive on agents in these threads are talking about two things: current, existing tools, and the existing rate of progress. Check out e.g. Epoch.ai for great industry analyses. To compare AI to crypto is disingenuous, they are completely different and crypto is a technology that fundamentally makes no sense in a world where governments want to (and arguably should) control money supply. You may or may not agree on that take but AI is something that governments will push aggressively and see as crucial to national security/control. It means this is not going away
> I’ve done things with Claude I never thought possible for myself to do,
That's the point champ. They seem great to people when they apply them to some domain they are not competent it, that's because they cannot evaluate the issues. So you've never programmed but can now scaffold a React application and basic backend in a couple of hours? Good for you, but for the love of god have someone more experienced check it before you push into production. Once you apply them to any area where you have at least moderate competence, you will see all sorts of issues that you just cannot unsee. Security and performance is often an issue, not to mention the quality of code....
This is remarkably dismissive and comes across as arrogant. In reality they assist many people with expert skills in a domain in getting things done in areas they are competent in, without getting bogged down in tedium.
They need a heavy hand to police to make sure they do the right thing. Garbage in, garbage out.
The smarter the hand of the person driving them, the better the output. You see a problem, you correct it. Or make them correct it. The stronger the foundation they're starting from, the better the production.
It's basically the opposite of what you're asserting here.
> So you've never programmed but can now scaffold a React application and basic backend in a couple of hours?
Ahaha, weren’t you the guy who wrote an opus about planes? Is this your baseline for “stuff where LLMs break and real engineering comes into the room”? There’s a harsh wake up call for you around the corner.
What wake up call mate? I've been on board as early adopter with GH Copilot closed beta since 2021, it was around time when you did not even hear about the LLMs. I am just being realistic about the limits of the technology. In the 90s, we did not need to convince people about the Internet. It just worked. Also - what opus? Have the LLMs affected your attention span so much, that you consider what typically an primary school first-grader would read during their first school class, an "opus" no less? No wonder you are so easily impressed.
I expect it’s your “I’m an expert and everyone else is merely an idiot child” attitude that’s probably making it hard to take you seriously.
And don’t get me wrong - I totally understand this personality. There are a similar few I’ve worked with recently who are broadly quite skeptical of what seems to be an obvious fact to me - their roles will need to change and their skillsets will have to develop to take advantage of this new technology.
I am a bit tired of explaining, but I run my own company, so its not like I have to fear my "roles and responsibilities" changing - I am designing them myself. I also am not a general skeptic of the "YAGNI" type - my company and myself have been early adopters on many trends. Those that made sense of course. We also tried to be early adopters of LLMs, all the way since 2021. And I am sorry if that sounds arrogant to you, but anyone still working on them and with them to me looks like the folks who were trying to build computers and TVs with the vaccuum tubes. With the difference that vaccuum tubes computers were actually useful at the time.
95% of companies fail. Yours will too, don't worry. Amazon themselves have already been using in-house versions of this to build AWS for over a year https://kiro.dev/ you can either continue adopting AI in your company or you can start filing your company bankruptcy papers
What would you need to see to change your mind? I can generate at mind-boggling scale. What’s your threshold for realizing you might not have explored every possible vector for AI capabilities?
What you wrote here was relevant about 9 months ago. It’s now outdated. The pace and velocity of improvement of Ai can only be described as violent. It is so fast that there are many people like you who don’t get it.
The last big release from OpenAI was a big giant billion-dollar flop. Its lackluster update was written about far and wide, even here on HN. But maybe you're living in an alternate reality?
My experience comes from the fact that after over a decade of working as a swe I no longer write code. It’s not some alternate reality thing or reading headlines. It’s my daily life that has changed.
Have you used AI before? Agentic systems are set up so it gives you a diff before even making committing to a change. Sounds like you haven’t really used AI agentically yet.
Disrespect the trend line and get rolled over by the steamroller. Labs are cooking and what is available commercially is lobotomized for safety and alignment. If your baseline of current max capability is sonnet 4.5 released just this summer you’re going to be very surprised in the next few months.
Right, like I was steamrolled by the "Team of Pocket Ph.D Experts" announced earlier this year with ChatGPT 5 ? Remember that underwhelming experience? The Grok to which you could "paste your entire source code file"? The constantly debilitating Claude models? Satya Nadella desperately dropping down to a PO role and bypassing his executives to try and micro-manage Copilot product development because the O365 Copilot experience is experiencing a MASSIVE pushback globally from teams and companies forced to use it ? Or is there another steamrolling coming around? What is this time? Zuckerberg implements 3D avatars in a metaverse with legs that can walk around and talk to us via LLMs? And then they sit down at virtual desks and type on virtual keyboards to produce software? Enlighten me please!
First examine your post. Can you create a 3D avatar with legs that can walk and talk?
If not, then for this area you’ve been steam rolled.
Anyway main point is, you’re looking at the hype headlines which are ludicrous. Where most optimists come from is that they are using it in the daily to code. To them it’s right in front of their eyes.
I’m not sure what your experience is but my opinion on AI doesn’t come from speculation. It comes from on the ground experience on how AI currently has changed my job role completely. If I hold the technology to be fixed and to not improve into the future then my point still stands. I’m not speculating. Most AI optimists aren’t speculating.
The current on the ground performance is what’s causing the divide. Some people have seen it fully others only have a rudimentary trial.
I have a hard time trusting the judgement of someone writing this:
> I no longer write code. I’ve been a swe for over a decade. AI writes all my code following my instructions. My code output is now expected to be 5x what it was before because we are now augmented by AI. All my coworkers use AI. We don’t use ChatGPT we use anthropic. If I didn’t use AI I would be fired for being too slow.
You should drop the prejudice and focus to be aware of the situation. This is happening all over the world, most people who have crossed this bridge just don’t share, just like they don’t share that they’ve brushed their teeth this morning.
People are sharing it. Look at this entire thread. It’s so conflicted.
We have half the thread saying it’s 5x and the other half saying they’re delusional and lack critical thinking.
I think it’s obvious who lacks critical thinking. If half the thread is saying on the ground AI has changed things and the other half just labels everyone as crazy without investigation… guess which one didn’t do any critical thinking?
Last week I built an app that cross compiled into Tauri and electron that’s essentially a google earth clone for farms. It uses mapbox and deckgl and you can play back gps tracks of tractor movements and the gps traces change color as the tractor moves in actual real time. There’s pausing, seeking, bookmarking, skipping. All happening in real time because it’s optimized to use shader code and uniforms to do all these updates rather than redrawing the layers. There’s also color grading for gps fix values and satellite counts which the user can switch instantaneously to with zero slow down on tracks with thousands and thousands of points. It all interfaces with an API that scans gcp storage for gps tracks and organizes it into a queryable api that interfaces with our firebase based authentication. The backend is deployed by terraform and written in strictly typed typescript and it’s automatically deployed and checked by GHA. Of course the electron and tauri app have GUI login interfaces that work fully correctly with the backend api and it all looks professionally designed like a movie player merged with Google earth for farm orchards.
I have rudimentary understanding for many of the technologies involved in the above. But I was able to write that whole internal tool in less than a week thanks to AI. I couldn’t have pulled it off without rudimentary understanding of the tech so some novice swe couldn’t really do it without the optimizations I used but that’s literally all I needed. I never wrote shader code for prod in my life and left to its own devices the AI would have come up with an implementation that’s too laggy to work properly.
That’s all that’s needed. Some basic high level understanding and AI did everything else and now our company has an internal tool that is polished beyond anything that would’ve been given effort to before AI.
I’m willing to bet you didn’t use AI agents in a meaningful way. Maybe copying and pasting some snippets of code into a chatbot and not liking the output. And then you do it every couple of weeks to have your finger on the pulse of AI.
Go deeper. Build an app with AI. Hand hold it into building something you never built before. It’s essentially a pair programming endeavor. Im willing to bet you haven’t done this. Go in with the goal of building something polished and don’t automatically dismiss it when the AI does something stupid (it inevitably will) Doing this is what actual “critical thinking” is.
> I think it’s obvious who lacks critical thinking.
My critical thinking is sharp enough to recognize that you're the recently banned
ninetyninenine user [0]. Just as unbalanced and quarrelsome as before I can see. It's probably better to draw some conclusion from a ban and adjust, or just leave.
Bro no one said 5x now or your fired that’s your own imagination adding flavor to it.
It’s obvious to anyone if your output is 5x less than everyone else you will eventually be let go. There’s no paradigm shift where the boss suddenly announced that. But the underlying unsaid expectation is obvious given what everyone is doing.
What happened was this, a couple new hires and some current employees started were using AI. There output was magnified and they were not only having more output but they were deploying code outside their areas of expertise doing dev ops, infra, backend and frontend.
This spread and within months everyone in the company was doing it. The boss can now throw a frontend job to a backend developer and now expect completion in a day or less. This isn’t every task but such output for the majority of tasks it’s normal.
If you’re not meeting that norm it’s blindingly obvious. The boss doesn’t need to announce anything when everyone is faster. There was no deliberate culture shift where the boss announced it. The closest equivalent is the boss hiring a 10x engineer to work alongside you and you have to scramble to catch up. The difference is now we know exactly what is making each engineer 10x and we can use that tool to also operate at that level.
Critical thinking my ass. You’re just labeling and assuming things with your premeditated subconscious bias. If anything it’s your perspective that is religious.
Also can you please stop stalking me and just respond to my points instead of digging through my whole profile and attempting to do character assassinations based off of what I wrote in the past? Thanks.
Whether you agree with it or not is besides the point. The point is it’s happening.
Your initial stance was disbelief. Now you’re just looking down at it as unprofessional.
Bro, I fucking agree. It’s unprofessional. But the entire point initially was that you didn’t believe it and my objective was to tell you that this is what’s happening in reality. Scoff at it all you want, as AI improves less and less “professional” people will be able to enter our field and operate at the same level as us.
I don't understand this idea that non-believers will be "steamrolled" by those who are currently adopting AI into their workflows. If their claims are validated and the new AI workflows end up achieving that claimed 10x productivity speedup, or even a 2x speedup, nobody is cursed to be steamrolled - they'll simply adopt those same workflows same as everyone else. In the meantime they aren't wasting their time trying to figure out the best way to coax and beg the LLM's into better performance.
That's actually what I'm arguing for; use tools where they are applicable. I'm against blind contrarianism and the 'nothing ever happens' attitude since that IME is being proven more wrong each week.
Seems fine, works, is fine, is better than if you had me go off and write it on my own. You realize you can check the results? You can use Claude to help you understand the changes as you read through them? I mean I just don’t get this weird “it makes mistakes and it’s horrible if you understand the domain that it is generating over” I mean yes definitely sometimes and definitely not other times. What happens if I DONT have someone more experienced to consult with or that will ignore me because they are busy or be wrong because they are also imperfect and not focused. It’s really hard to be convinced that this point of view is not just some knee jerk reaction justified post hoc
Yes you can ask them "to check it for you". The only little problem is as you said yourself "they make mistakes", therefore : YOU CANNOT TRUST THEM. Just because you tell them to "check it" does not mean they will get it right this time. Again, however it seems "fine" to you, please, please, please / have a more senior person check that crap before you inflict serious damage somewhere.
Nope, you read their code, ask them to summarize changes to guide your reading, ask it why it made certain decisions you don’t understand and if you don’t like their explanations you change it (with the agent!). Own and be responsible for the code you commit. I am the “most senior”, and at large tech companies that track, higher level IC corresponds to more AI usage, hmm almost like it’s a useful tool.
Ok but you understand that the fundamental nature of LLMs amplifies errors, right? A hallucination is, by definition, a series of tokens which is plausible enough to be indistinguishable from fact to the model. If you ask an LLM to explain its own hallucinations to you, it will gladly do so, and do it in a way that makes them seem utterly natural. If you ask an LLM to explain its motivations for having done something, it will extemporize whichever motivation feels the most plausible in the moment you're asking it.
LLMs can be handy, but they're not trustworthy. "Own and be responsible for the code you commit" is an impossible ideal to uphold if you never actually sit down and internalize the code in your code base. No "summaries," no "explanations."
So your argument is that if people don't use the tool correctly they might get incorrect results? How is that relevant? If you Google search for the wrong query you'll similarly get incorrect results
I would say while LLMs do improve productivity sometimes, I have to say I flatly cannot believe a claim (at least without direct demonstration or evidence) that one person is doing the work of 20 with them in december 2025 at least.
I mean from the off, people were claiming 10x probably mostly because it's a nice round number, but those claims quickly fell out of the mainstream as people realised it's just not that big a multiplier in practice in the real world.
I don't think we're seeing this in the market, anywhere. Something like 1 engineer doing the job of 20, what you're talking about is basically whole departments at mid sized companies compressing to one person. Think about that, that has implications for all the additional management staff on top of the 20 engineers too.
It'd either be a complete restructure and rethink of the way software orgs work, or we'd be seeing just incredible, crazy deltas in output of software companies this year of the type that couldn't be ignored, they'd be impossible to not notice.
This is just plainly not happening. Look, if it happens, it happens, 26, 27, 28 or 38. It'll be a cool and interesting new world if it does. But it's just... not happened or happening in 25.
I would say it varies from 0x to a modest 2x. It can help you write good code quickly, but, I only spent about 20-30% of my time writing code anyway before AI. It definitely makes debugging and research tasks much easier as well. I would confidently say my job as a senior dev has gotten a lot easier and less stressful as a result of these tools.
One other thing I have seen however is the 0x case, where you have given too much control to the llm, it codes both you and itself into pan’s labyrinth, and you end up having to take a weed wacker to the whole project or start from scratch.
That's why you dont use LLMs as a knowledge source without giving them tools.
"Agents use tools in a loop to achieve a goal."
If you don't give any tools, you get hallucinations and half-truths.
But you give one a tool to do, say, web searches and it's going to be a lot smarter. That's where 90% of the innovation with "AI" today is coming from. The raw models aren't gettin that much smarter anymore, but the scaffolding and frameworks around them are.
Tools are the main reason Claude Code is as good as it is compared to the competition.
> The raw models aren't gettin that much smarter anymore, but the scaffolding and frameworks around them are.
yes, that is my understanding as well, though it gets me thinking if that is true, then what real value is the llm on the server compared to doing that locally + tools?
You still can't beat an acre of specialized compute with any kind of home hardware. That's pretty much the power of cloud LLMs.
For a tool use loop local models are getting to "OK" levels, when they get to "pretty good", most of my own stuff can run locally, basically just coordinating tool calls.
Of course, step one is always critically think and evaluate for bad information. I think for research, I mainly use it for things that are testable/verifiable, for example I used it for a tricky proxy chain set up. I did try to use it to learn a language a few months ago which I think was counter productive for the reasons you mentioned.
How can you critically assess something in a field you're not already an expert on?
That Python you just got might look good, but could be rewritten from 50 lines to 5, it's written in 2010-style, it's not using modern libraries, it's not using modern syntax.
And it is 50 to 5. That is the scale we're talking about in a good 75% of AI produced code unless you challenge it constantly. Not using modern syntax to reduce boilerplate, over-guarding against impossible state, ridiculous amounts of error handling. It is basically a junior dev on steriods.
Most of the time you have no idea that most of that code is totally unnecessary unless you're already an expert in that language AND libraries it's using. And you're rarely an expert in both or you wouldn't even be asking as it would have been quicker to write the code than even write the prompt for the AI.
I use web search (DDG) and I don’t think I have ever try more than one queries in the vast majority of cases. Why because I know where the answer is, I’m using the search engine as an index to where I can find it. Like “csv python” to find that page in the doc.
It's entirely dependent on the type of code being written. For verbose, straightforward code with clear cut test scenarios, one agent can easily 24/7 the work of 20 FT engineers. This is a best case scenario.
Your productivity boost will depend entirely on a combination of how much you can remove yourself from the loop (basically, the cost of validation per turn) and how amenable the task/your code is to agents (which determines your P(success)).
Low P(success) isn't a problem if there's no engineer time cost to validation, the agent can just grind the problem out in the background, and obviously if P(success) is high the cost of validation isn't a big deal. The productivity killer is when P(success) is low and the cost of validation is high, these circumstances can push you into the red with agents very quickly.
Thus the key to agents being a force multiplier is to focus on reducing validation costs, increasing P(success) and developing intuition relating to when to back off on pulling the slot machine in favor of more research. This is assuming you're speccing out what you're building so the agent doesn't make poor architectural/algorithmic choices that hamstring you down the line.
Respectfully, if
I may offer constructive criticism, I’d hope this isn’t how you communicate to software developers, customers, prospects, or fellow entrepreneurs.
To be direct, this reads like a fluff comment written by AI with an emphasis on probability and metrics. P(that) || that.
I’ve written software used by a local real estate company to the Mars Perseverance rover. AI is a phenomenally useful tool. But be weary of preposterous claims.
I'll take you at your word regarding respectfully. That was an off the cuff attempt to explain the real levers that control the viability of agents under particular circumstances. The target market wasn't your average business potato but someone who might care about a hand waived "order approximate" estimator kind of like big-O notation, which is equally hand waivey.
Given that, if you want to revisit your comment in a constructive way rather than doing an empty drive by, I'll read your words with an open mind.
> It's entirely dependent on the type of code being written. For verbose, straightforward code with clear cut test scenarios, one agent can easily 24/7 the work of 20 FT engineers. This is a best case scenario.
So the "verbose, straightforward code with clear cut test scenarios" is already written by a human?
> I mean from the off, people were claiming 10x probably mostly because it's a nice round number,
Purely anecdotal, but I've seen that level of productivity from the vibe tools we have in my workplace.
The main issue is that 1 engineer needs to have the skills of those 20 engineers so they can see where the vibe coding has gone wrong. Without that it falls apart.
My experience is that you get out what you put in. If you have a well-defined foundation, AI can populate the stubs and get it 95% correct. Getting to that point can take a bit of thought, and AI can help with that, too, but if you lean on it too much, you'll get a mess.
And of course, getting to the point where you can write a good foundation has always been the bulk of the work. I don't see that changing anytime soon.
This is completely wrong. Codex 5.2 and Claude Sonnet 4.5 don't have any of these issues. They will regularly tell you that you're wrong if you bother to ask them and they will explain why and what a better solution is. They don't make up anything. The code they produce is noticeably more efficient in LoC than previous models. And yes they really will do research, they will search the Internet for docs and articles as needed and cite their references inline with their answers.
You talk as if you haven't used a LLM since 2024. It's now almost 2026 and things have changed a lot.
With apologies, and not GP, but this has been the same feedback I've personally seen on every single model release.
Whenever I discuss the problems that my peers and I have using these things, it's always something along the lines of "but model X.Y solves all that!", so I obediently try again, waste a huge amount of time, and come back to the conclusion that these things aren't great at generation, but they are fantastic at summarization and classification.
When I use them for those tasks, they have real value. For creation? Not so much.
I've stopped getting excited about the "but model X.Y!!" thing. Maybe they are improving? I just personally haven't seen it.
But according to the AI hypers, just like with every other tech hype that's died over the past 30 years, "I must just be doing it wrong".
A lot of people are consistently getting their low expectations disproven when it comes to progress in AI tooling. If you read back in my comment history, six months ago I was posting about how AI is over hyped BS. But I kept using it and eventually new releases of models and tools solved most of the problems I had with them. If it has not happened for you yet then I expect it will eventually. Keep up with using the tools and models and follow their advancements and I think you'll eventually get to the point where your needs are met
I'd be willing to give you access to the experiment I mentioned in a separate reply (have a github repo), as far as the output that you can get for a complex app buildout.
Will admit It's not great (probably not even good) but it definitely has throughput despite my absolute lack of caring that much [0]. Once I get past a certain stage I am thinking of doing an A-B test where I take an earlier commit and try again while paying more attention... (But I at least want to get where there is a full suite of UOW cases before I do that, for comparison's sake.)
> Those twenty engineers must not have produced much.
I've been considered a 'very fast' engineer at most shops (e.x. at multiple shops, stories assigned to me would have a <1 multiplier for points[1])
20 is a bit bloated, unless we are talking about WITCH tier. I definitely can get done in 2-3 hours what could take me a day. I say it that way because at best it's 1-2 hours but other times it's longer, some folks remember the 'best' rather than median.
[0] - It started as 'prompt only', although after a certain point I did start being more aggressive with personal edits.
[1] - IDK why they did it that way instead of capacity, OTOH that saved me when it came to being assigned Manual Testing stories...
It's not 20 and it's not 2. It's not a person. It's a tool. It can make a person 100x more effective at certain specific things. It can make them 50% less effective at other things. I think, for most people and most things, it might be like a 25% performance boost, amortized over all (impactful) projects and time, but nobody can hope to quantify that with any degree of credibility yet.
> but nobody can hope to quantify that with any degree of credibility yet
i'd like to think if it was really good, we would see product quality improve over time; iow less reported bugs, less support incidents, increased sign-ups etc, that could easily be quantified no?
Yeah, it makes me wonder whether I should start learning to be a carpenter or something. Those who either support AI or thinks "it's all bullshit" cite a lack of evidence for humans truly being replaced in the engineering process, but that's just the thing; the unprecedented levels of uncertainty make it very difficult to invest one's self in the present, intellectually and emotionally. With the current state of things, I don't think it's silly to wonder "what's the point" if another 5 years of this trajectory is going to mean not getting hired as a software dev again unless you have a PhD and want to work for an AI company.
What doesn't help is that the current state of AI adoption is heavily top-down. What I mean is the buy-in is coming from the leadership class and the shareholder class, both of whom have the incentive to remove the necessary evil of human beings from their processes. Ironically, these classes are perhaps the least qualified to decide whether generative AI can replace swathes of their workforce without serious unforeseen consequences. To make matters worse, those consequences might be as distal as too many NEETs in the system such that no one can afford to buy their crap anymore; good luck getting anyone focused on making it to the next financial quarter to give a shit about that. And that's really all that matters at the end of the day; what leadership believes, whether or not they are in touch with reality.
His logic is off and his experience is irrelevant because i doesn’t encompass scale to have been exposed to an actual paradigm shifting event. Civilizations and entire technologies have been overturned so he can’t say it won’t happen this time.
What we do know is this. If AI keeps improving at the current rate it’s improving then it will eventually hit a point where we don’t need software engineers. That’s inevitable. The way for it to not happen is for this technology to hit an impenetrable wall.
This wave of AI came so fast that there are still stubborn people who think it’s a stochastic parrot. They missed the boat.
It’s strange that the article says the white collar worker in nyc and small business owner in suburban Texas are not the same market. To many businesses they are in the same market. McDonald’s Home Depot etc they don’t make different products for those two individuals
Author here. I think this thread is mixing two very different kinds of markets, so let me clarify the scope of the argument.
I agree with the point that markets are often defined by legal and operational systems — how contracts work, how labor is regulated, how payments and compliance function. That’s exactly why country or jurisdiction boundaries sometimes matter a lot.
Where I think we’re talking past each other is the Home Depot / McDonald’s examples.
Those are low-involvement, highly standardized, commodity-style businesses. Their products, pricing logic, and purchasing situations are intentionally broad. In that world, a white-collar worker in NYC and a small business owner in suburban Texas can absolutely be treated as “the same market” for many decisions, because the offer is designed to ignore sharp differences.
The article isn’t arguing against that. It’s explicitly about sharper products — especially startups, B2B tools, workflow software, education, compliance-heavy or behavior-changing products — where the purchasing situation narrows quickly.
In those cases, what matters isn’t whether two people can physically buy the same thing, but whether the same offer survives the same constraints and produces the same outcome. Authority to buy, risk tolerance, institutional expectations, and default alternatives diverge much faster there, even within the same legal system.
So yes, commodity retail is a valid counterexample — but it’s also a special case. The failure pattern the article is pointing at shows up when teams implicitly assume their product behaves like a Big Mac or a box of nails, when in reality it behaves more like a change in how work, learning, or decision-making happens.
That mismatch is where “same country = same market” becomes dangerous
home depot doesn't have lucatations in Mahattan - I don't even need to check that made up fact to believe it. The market in manhattan cannot support home depot as it opperates in the Texas suburb. Even if they do happen to have a store there it would have to be different.
I'm just enjoying the last few years of this career. Let me have fun!
Joking aside, we have to understand that this is the way software is being created and this tool is going to be the tool most trivial software (which most of us make) will be created with.
I feel like the industry is telling me: Adopt of become irrelevant
Meh, I am also old enough to have experienced what the GP post mentioned, and I remember also when Visual Basic 6 was released, a similar sentiment appeared:
Suddenly, every cousin 13 year old could implement apps for their Uncle's dental office, laboratory, parts shop billing, tourism office management, etc. Some people also believed that software developers would become irrelevant in couple of years.
For me as an old programmer, I am having A BLAST using these tools. I have used enough tools (TurboBasic, Rational Rose (model based development, ha!), NetBeans, Eclipse, VB6, BorlandC++ builder) to be able to identify their limits and work with them.
That's great! I am also having a blast, and trying hard to take advantage of the new capability while not turning into the programmer equivalent of the passengers of the BNL Axiom. We aren't the intended audience for this post.
Since they are not showing you how this model compares against the benchmarks they are showing, here is a quick view with the public numbers from Google and Anthropic. At least this gives some context:
SWE-Bench (Pro / Verified)
Model | Pro (%) | Verified (%)
--------------------+---------+--------------
GPT-5.2-Codex | 56.4 | ~80
GPT-5.2 | 55.6 | ~80
Claude Opus 4.5 | n/a | ~80.9
Gemini 3 Pro | n/a | ~76.2
And for terminal workflows, where agentic steps matter:
Terminal-Bench 2.0
Model | Score (%)
--------------------+-----------
Claude Opus 4.5 | ~60+
Gemini 3 Pro | ~54
GPT-5.2-Codex | ~47
So yes, GPT-5.2-Codex is good, but when you put it next to its real competitors:
- Claude is still ahead on strict coding + terminal-style tasks
- Gemini is better for huge context + multimodal reasoning
- GPT-5.2-Codex is strong but not clearly the new state of the art across the board
It feels a bit odd that the page only shows internal numbers instead of placing them next to the other leaders.
I like Opus 4.5 a lot, but a general comment on benchmarks: the number of subtasks or problems in each one is finite, and many of the benchmarks are saturating, so the effective number of problems at the frontier is even smaller. If you think of the generalizable capability of the model as a latent feature to be measured by benchmarks, we therefore have only rather noisy estimates. People read too much into small differences in numbers. It's best to aggregate across many, Epoch has their Capabilities Index, and Artificial Analysis is doing something similar, and probably others I don't know or remember.
And then there's the part of models that is hard to measure. Opus has some sort of HAL-like smoothness I don't see in other models, but meanwhile, I haven't tried gpt-5.2 for coding yet. (Neither Gemini 3 Pro; I'm not claiming superiority of Opus, just that something in practical usability is hard to measure.)
I'm finding that the newer GPT models are much more willing to leverage tools/skills than Claude, reducing interventions requesting approval. Just an observation.
reply