Ask HN: What "developer holy war" have you flip-flopped on?
10 points by meowface 1d ago 31 comments
Ask HN: How do you connect with other founders in your city?
5 points by leonagano 1d ago 2 comments
Anthropic's CEO says in 3-6 months, AI will write 90% of the code (March 2025)
36 amarcheschi 76 8/16/2025, 9:50:24 AM businessinsider.com ↗
Or in other words, of course he says that; it's not really a useful analysis is it?
And UML even made a lot more sense to precisely describe a problem compared to human language prompts.
IME it's more useful to treat a source file (or new project) like a sketchbook, you start with an empty sheet of paper but already have a rough idea what you want, then you quickly sketch out the outlines, try out different ideas, explore different solutions (all without going into too much detail yet), you step through the code, explore some different paths to get an idea what "feels right", delete things, shuffle them around, rewrite them and slowly filling out the details and that way incrementally get towards the first working version.
Which such an approach of incremental "micro-feedback-loops" you already eliminated a lot of dead ends that will appear anyway despite the best plans - but identifying such dead ends early is much better than late.
This incremental approach also forces you to keep the code small, tidy and malleable. The planning stage basically already happens in source code, and it's not one long planning stage, but many micro-planning-stages.
...this is also my problem with the current state of LLMs, they're pretty good at creating a first initial sketch from a very general problem statement (but only for things that have been done thousands of times by other people - but that's a different issue...) - but the more you'll need to go into the details and discover and fix problems in your initial 'mental design' (e.g. discovering what you actually wanted in the first place), the more the differences to traditional programming disappear, and at some tipping point it gets even more complex because you need to steer the LLM with a language (the human language) that simply isn't useful enough for detailed problem descriptions - that's why lawyers, mathematicians, engineers and scientists all invented their own precise 'DSLs'.
After a few iterations your prompts need to be just as detailed as writing source code in the first place - so what's the point of writing prompts in human language again?
I feel like productivity could be improved much more by improving programming tools (yeah - boring old-school incremental maintenance work) instead of betting on some weird AI future which will just move the focus from writing source code to writing human-language prompts (which IMHO is a definitive step backwards because human language lacks precision - and adding that precision is how you end up with programming languages).
Why are editing, compilation and debugging/testing still separate steps, why do debuggers still have those 60s style bare variable panels instead of realtime visualizations of the internal program state? Why do I still need to wait for a build to finish? Why is version control still such a PITA to work with? Instead everybody is jumping on the AI hype train while everything around them is crumbling into a post-apocalyptic wasteland.
A Driverless Tesla Will Travel From L.A. to NYC by 2017, Says Musk
https://www.nbcnews.com/business/autos/driverless-tesla-will...
Ray Kurzweil was even more on the money in his 2006 book "The singularity is near". I remember reading some of the stuff in there that is now happening. From the top of my head he predicted 2030 for human-level AI hardware and software that could be bought for 1000 USD by anyone. I feel he's going to be very close.
No comments yet
https://www.techradar.com/pro/nearly-half-of-all-code-genera...
I've also worked with juniors who are technically much better than most of the senior people around them.
Give me them downvotes all you want but I’m not the “senior” turning in slop Llm code.
[1] https://en.wikipedia.org/wiki/Bullshit_Jobs
Apparently it's lots of fiduciary duties.
As with driving cars, even if the AI is strictly better at doing these tasks than they are at writing code, mistakes aren't so easy to recover from and can destroy something unrecoverably in a 5-second attention lapse from a human overseer.
Lots of software has a test environment. Even in live, e.g. bank account losing your money the transactions can be un-wound.
And that's the difference when it comes to replacing software devs with LLMs vs replacing CEOs with LLMs: it's possible to write the tests and then lock them. And to perform code review before changes are merged into the main branch. And to test against something other than production.
I know the Board can in principle remove a CEO, but is there even a standardised way for a CEO to have a vice-CEO that always checks everything they do, that always tests their commands against a simulation of the company and only lets them proceed if they agree with the outcome?
The point is that "AI as CEO" would be in the category of "business-critical" software, and also that current approaches to AI also lack sufficient guarantees of obligation compliance or sufficient defence against failures, which in the banking example would be things like the AI deciding to save money by deleting the system capable of unwinding incorrect banking transactions.
To the extent this kind of failure mode happens with vibe coding (in the original coining of the term: always accept without reading), it's like letting the LLM modify the unit tests to always succeed regardless of the code.
But I'm not expecting that to hold true for linux kernel or postgres codebase or equivalent anytime soon.
With every piece of code created i write a doc of what id need to focus on when rewriting chunks.
AI has its usage, it makes some things faster... but spitting out more code does not necessarily make it more productive.
I ride AI train, but i ride my cb650r same way... sometimes wondering why i am still alive.
For my latest project, I have 20k+ lines of markdown docs to guide it, with great success. Some of them are generic rules, some of them describes how I code, some of them describing the codebase & features. Then I have another 17k+ that are used while coding. Plans, phases, todos, reviews etc.
All of them are written by Claude Code also. I'm calling it "spec-driven development".
Cursor has a much different flow, where you usually pair program with it, which I call "ai-assisted development".
My current approach is to stub out the architecture I want, write the type definitions I want in full, and ask Claude to fill in the blanks. It's good at implementing an interface but bad at devising the interface on its own.
It’s not coding in English, but giving Claude memory about how the code base works, how it’s structured, what features available, why things are created etc.
This is how we work too, we have all these info in our head when building. But every CC session starts a new with blank slate. These documents helps getting the session “warmed up” enough to get quality output with lesser tech debt.
Source: my commits pass reviews and it's not because the reviewers don't have like 50 years of experience between them.
It's funny how the comments on HN went from complete denial about coding with LLM's ever being possible at all like 1 year ago to "oh, it might be happening" now. I find it hard to understand how people can simultaneously be skilled (older) engineers while at the same time being completely oblivious to the concept of exponential improvement of tech in general. I guess people don't WANT to see the reality that is change and prefer to just look away and cope?
Yes. But frankly, it's hard to blame them. It's just a shitty social contract where you spend lots of time gaining skills and by the time you're done your field has been completely obsoleted.
I'm a huge fan of making comparisons to how things used to work before industrial revolution, because that's what our brains evolved for. If your father was a blacksmith, you'd be a blacksmith too, and so would be your son. It was difficult to imagine a world that simply would stop needing blacksmiths. So it's completely normal that our brains spend first 20 years learning, and once we're done learning, we just apply that knowledge to everyday life, save for minor tweaks here and there.
Then it became a requirement to keep learning your entire life because the world kept changing. This sucks, but it's not completely infeasible, so people rolled with it, especially that new positions were cozy and well-paid, and there was some gentle exit path for those who were too old to learn new things.
But now we're in a situation where the society tells you that there's huge demand for given skill, and then your entire profession gets completely obliberated before you're even halfway done with your studies. In such environment it's simply impossible to create a realistic career plan, let alone have energy for luxuries such as raising a family or maintaining social connections. This isn't good.
AI has objectively made our life worse, and everyone is completely clueless how the situation develops. Denying everything and hoping to wait things out might actually be just as good of an attitude like any other. Because honestly tell me - even if we assume the most optimistic scenario "AGI within five years" - how the fuck do I plan my career?
I will come back in 3-6 months and hope its better to understand its own limitations.
I'm guessing it would be less then a week for me to write it on my own.
The biggest issue is that I've now seen the shit it creates so I have zero trust in the code I now have from a security and stability standpoint. I know many have better experiences then mine.
The question is not how many characters can a computer spit out, but whether you need a human in the loop or not. That's not going to change in 6 months. In a few years, perhaps.
That being said, I had to be very detailed and meticulous because they go off the rails and make poor decisions
Just means I type and search a lot less, and can get things out of my head and into computers much faster
Project managers be like sending me AI code snippets like they know shit.
Exactly when depends what you count, but part of the path leading to ChatGPT was getting a bunch of dumb systems to work together to train a smarter system: https://arxiv.org/pdf/1909.08593 fig 1, and also section 4.4 for what can go wrong
Not sure either outcome is likely though.
Someone also said extending this pattern into the future is probably why computers in Star Trek mess up
Working on undocumented legacy code (read: probably the majority of all embedded code in the world) is now like 100x faster. No need to plow through a massive codebase and trying to grasp expired toolchains for weeks/months anymore.
How is that not part of programming?
Or another way, if Rust didn't exist, could an AI create Rust?
https://www.businessinsider.com/github-ceo-developers-embrac...
https://www.theverge.com/news/757461/microsoft-github-thomas...
This industry is full of snake oil salesmen and false promises, it's reliant on hype and an army of fanboys trained on twitter to fan the flames and create the impression of a revolution that isn't there and might never be there, just to prop the valuations of the likes of OpenAI and Anthropic.
It always follows the same playbook:
- The guru type e.g muskie, sama, adam neumann, elizabeth holmes, SBF, says something on twitter: "XYZ is the future"
- Army of same-thinkers lurkers start getting excited about it.
- VC fund bros who are also forming their opinion on vibes stalking twitter, also get excited about it.
- Big money poured in XYZ thing.
- Guru keeps claiming that XYZ version 2.0 is just behind the corner (AGI/Superintelligence, Autonomous Self Driving Cars, the Boring Tunnel, cure for whatever) projecting insane and out of touch with reality timelines.
- XYZ thing either becomes the future (not likely), or gets abandoned (remember when buying pixelized pics of apes on the internet was an "investment"?) and the collective amnesia of the fanbase doesn't keep the guru accountable. It's easy to make hundreds of claims if getting it right once, even by chance, makes you some kind of a prophet and all your wrong claims of the past are dismissed due to your fanbase's dog-like adoration for you.
- Cycle repeats.
The LLM craze reminds me of the NFT, web3, crypto grift all over again, but this time with a product that at least has some value.
Tom Renner wrote an excellent article on the "LLM inevitabilism" https://tomrenner.com/posts/llm-inevitabilism/
Now, AI is the new outsourcing it seems. Let's lower the bars of quality, see how much crappiness and glitchiness can be tolerated, freeze it right there as the new "industry standard" and move on to the next cost cutting thing.