Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The last six months in LLMs in five minutes

simonwillison.net

Post reply on HN
Print view View original

471–480 of 631 posts

Re: The last six months in LLMs in five minutes

#471

Post by nijave »

DennisP wrote: post #458
hollowturtle wrote: post #209

> The coding agents got really good It's since november 2025, the so called "inflection point", that I'm still wondering for who coding agents become "really good". All I observe they got better at tool call and answering questions about big codebases, especially if the question has a vague pattern to search, and they're superuseful for that! For generating production code even with a lot of steering and baby sitting…

Steve Yegge wrote about this in his book Vibe Coding . He says it takes about a year of experience before you're consistently getting good results. He writes about lots of different techniques for doing that, but also says a lot of it comes down to just getting a feel for when the LLM is going to go haywire.

I'd say closer to 6 months for me but probably still some room to improve.

I think getting a decent setup with a fast feedback loop for the agent combined with context (in repo markdown)+memories goes a long way.

After having Claude Code "remember" my preferences and tools, it's more efficient.

It has a tendency to copy existing patterns so a good AGENTS.md with best practices and architectural goals goes a long way to prevent it from duplicating patterns you're trying to get rid of.

Re: The last six months in LLMs in five minutes

#472

Post by chasd00 »

Glohrischi wrote: post #436
hiw2d wrote: post #430

Earlier quoted context omitted.

I dont have PMs or POs in my firm fella. Im hyper efficient. You clearly are not and are full of it. If youre only doing 5%, you should only get paid for that. lol. Are you happy to take a salary drop?

Whats wrong with you? Why the change of tone? I'm a Cloud ARchitect with experience in coding (15 years) and infrastructure (10 years) and startup founder... If you don't comprehend what i write, feel free to ask but don't be dick?

the change in tone is because you're either dealing with a troll or the CEO of a one person, contract-for-hire, "company".

Re: The last six months in LLMs in five minutes

#473

Post by nijave »

GistNoesis wrote: post #285

Last 6 months is humanity losing control of LLMs. - Memory market cornering which mitigated the adoption of local AI despite great open model being released. - Fast penetration of IP exfiltrating tools in companies world-wide. - Developers producing more code that they can read. - Autonomous agents killing Open Source by siphoning the attention economy - Autonomous agents destroyed online communities (including HN) -…

I think it's just further exposed cracks in software engineering that were always there.

Ideally we'll come out of the AI hype cycle having learned better practices.

Re: The last six months in LLMs in five minutes

#474

Post by nijave »

swed420 wrote: post #445
jimbobthemighty wrote: post #161

I asked Gemini for a video of 'pelican riding a unicycle in hyde park' - I was blown away by the output: https://gemini.google.com/share/55e250c99693

According to OP: > Why this test? Because pelicans are hard to draw, bicycles are hard to draw, pelicans can’t ride bicycles... and there’s zero chance any AI lab would train a model for such a ridiculous task. At this juncture I'm left wondering why competing AI labs wouldn't train for this now well known "test".

Given their proclivity to scrape the entire contents of the internet, it's only a matter of time intentional or otherwise.

I've heard the same has happened with common benchmarks (they've ingested solutions into training data)

Re: The last six months in LLMs in five minutes

#475

Post by nijave »

jimbobthemighty wrote: post #161

I asked Gemini for a video of 'pelican riding a unicycle in hyde park' - I was blown away by the output: https://gemini.google.com/share/55e250c99693

Google/Gemini has pretty impressive audio visual capabilities. I tried to have Claude add mulch to a landscape picture and it looked like someone hit it with the orange spray paint tool in MS Paint. Nano Banana actually produced something fairly realistic

Re: The last six months in LLMs in five minutes

#476

Post by noisy_boy »

DennisP wrote: post #458
hollowturtle wrote: post #209

> The coding agents got really good It's since november 2025, the so called "inflection point", that I'm still wondering for who coding agents become "really good". All I observe they got better at tool call and answering questions about big codebases, especially if the question has a vague pattern to search, and they're superuseful for that! For generating production code even with a lot of steering and baby sitting…

Steve Yegge wrote about this in his book Vibe Coding . He says it takes about a year of experience before you're consistently getting good results. He writes about lots of different techniques for doing that, but also says a lot of it comes down to just getting a feel for when the LLM is going to go haywire.

> but also says a lot of it comes down to just getting a feel for when the LLM is going to go haywire

That has been my experience too. The days when I'm very focused, being extra deliberate and constantly questioning/examining/challenging things, the results are much better. Autopilot days just go through in a daze and the outcome is objectively worse. This has made me much more hands-on and pushed me towards models which are actually not that "clever" like codex at effort=low but fast. Given that I'm doing the meat of the thinking, might as well not be slowed down by the model and lose the flow.

Re: The last six months in LLMs in five minutes

#477

Post by chasd00 »

hollowturtle wrote: post #462
epolanski wrote: post #418

Earlier quoted context omitted.

I've quoted you two tools (Ghostty and Redis) whose development now regularly uses AI assistance to deliver production code. I quoted those because their authors shared their experiences, the strengths and the limits of the tooling. There's many more, from Flask to Docker, from Ruby to FastAPI or Tanstack. LLVM has integrated AI-generated PRs, so did Swift and Mojo. Sasha Levin has pushed into Linux Nvidia-related ke…

You cited mostly web tech, which proves my point ;) Is antirez uses extensively agents to contribute to redis doesn't mean it's a becoming industry trend. I'd say quite the contrary, it isn't in the gaming industry for example, where novel ideas matter. And btw Antirez and Linus for example, put a lot of effort into steering agent into doing the right thing for them which is totally different than "these tools become…

GenAI is very much being used in the gaming industry, maybe its just you.

Re: The last six months in LLMs in five minutes

#478

Post by sjagauanbdvva »

kstenerud wrote: post #454
gilrain wrote: post #440

Earlier quoted context omitted.

> But I still stand by the quality of my code, including here. You and I don't need to agree. You aren’t having a disagreement with a person. You’re having a disagreement with reality.

> You aren’t having a disagreement with a person. You’re having a disagreement with reality. How so? Are you going to instruct us all on how a termcaps mismatch bug is an indicator of poor code quality, rather than an unfortunate bug emerging from within the chaos of the many layers of disparate technologies that must somehow be stitched together (along with their idiosyncrasies) in order to make a project like this…

Because you won’t listen to a word anyone says lol.

You had a visual bug right at the top of the repos README. Then insisted you hadn’t noticed it before.

Whats important is not that specific visual bug, it’s what that bug says about the rest of the code.

How can we believe that this code is high quality if we see a glaring issue 5 seconds into opening the github?

We didn’t seek out your repo and start lobbing critiques at it. YOU POSTED IT as an example of high quality generated code. I’m telling you I am unimpressed

Re: The last six months in LLMs in five minutes

#479

Post by rconti »

hollowturtle wrote: post #209

> The coding agents got really good It's since november 2025, the so called "inflection point", that I'm still wondering for who coding agents become "really good". All I observe they got better at tool call and answering questions about big codebases, especially if the question has a vague pattern to search, and they're superuseful for that! For generating production code even with a lot of steering and baby sitting…

I'm moderately horrified every time claude runs the same broken, YOLO SWAG git commands from stackoverflow, gets errors, tries a few more things, then finally figures out how to commit and push correctly.

Re: The last six months in LLMs in five minutes

#480

Post by keybored »

hiw2d wrote: post #407
Glohrischi wrote: post #402

Earlier quoted context omitted.

I had a really fun day yesterday because anthropics limits on their normal 20$ subscription allowed me to play around for the whole day without hitting a limit. Its 'production' code because its a small browser game which has very small to 0 requirements on security and being perfect but high requirements on 'ever even doing this' and 'fun'. The code it generated hat 0 compiletime errors. I was able to descripe 10 th…

"We could already be at 95% at 'ai will take your coding job' without knowing because these 5% are so relevant." This is nonsense. Im not a SWE but a CEO, if that were true I'd be firing without a hitch. And yet this is not the activity we see. Why is that? Perhaps merely writing code is not the entire job.

CEO makes fresh account to tell someone that writing code is not the entire job? I don’t buy it.
Post reply on HN