strangescript

Your assumption is you are "in control", as does everyone right before they have an accident.

you should probably avoid driving or riding in motor vehicles

I work 60+ hours a week with Claude Code CLI, always run dangerously skip, coding on multiple repos, on a mac. This has never happened. Nothing remotely close has ever happened. I have been using CC since research preview. I would love to know the series of prompts that lead to that moment.

My kids have book reports and stuff. Lately I can use AI to generate non-trivial questions about the books and use it to quiz them without me knowing anything about the books. Been super useful.

MS invested in Apple to keep them afloat so they had a competitor and avoid anti-trust lawsuits.

That seems more nefarious than buying products from each other. Either way it worked out fine for Apple.

Its so wild to me that people that should know better pretend that this kind of stuff doesn't happen in every industry.

LLMs write all my code now and I just have to review it. Not only has my output 3x'ed at least, I also have zero hesitations now tackling large refactors, or tracking down strange bugs. For example, I recently received a report there was some minor unicode related data corruption in some of our doc in our DBs. It was cosmetic, and low priority, also not a simple task to track down traditionally. But now I just put [llm agent on it, to avoid people accusing me of promoting] on it. It found 3 instances of the corruption across hundreds of documents and fixed them.

I am sure some of you are thinking "that is all slop code". It definitely can be if you don't do your due diligence in review. We have definitely seen a bifurcation of devs who do that, and those who don't, where I am currently working.

But by far the biggest gain is my mental battery is far less drained at the end of the day. No task feels soul crushing anymore.

Personally, coding agents are the greatest invention of my lifetime outside the emergence of the internet.

And by this time next year, this comment is going to look very silly

there are built in moderation tools you should turn on if you have external customers generating images, or inputing data that might be sketch

You didn't read the X replies if you believe that

Except they weren't intentionally trying to deceive anyone. They made the faulty assumption that these problems were non-trivial to solve and didn't think it was simply GPT-5 aggregating solutions in the wild.

This entire thing has been pretty disingenuous on both sides of the fence. All the anti-AI (or anti OpenAI) people are doing victory laps, but what GPT-5 Pro did is still very valuable.

1) What good is your open problem set if really its a trivial "google search" away from being solved. Why are they not catching any blame here?

2) These answers still weren't perfectly laid out for the most part. GPT-5 was still doing some cognitive lifting to piece it together.

If a human would have done this by hand it would have made news and instead the narrative would have been inverted to ask serious questions about the validity of some these style problem sets and/or ask the question how many other solutions are out there that just need pieced together from pre-existing research.

But, you know, AI Bad.

I love Karpathy, but he is wrong here. In a few short years we went from chat bots being toys and video creation predicted to be impossible in the near term to agents writing working apps and high def video that occasionally is indistinguishable from real life.

The rate depth, breadth and frequency of releases has only increased, not decreased. Meanwhile, everyone is waiting on bated breath for Gemini 3 to drop. A decade for reliable agents is not only comical, but willful cognitive dissonance at this point.

"grunt, gulp, webpack, coffeescript, babel" --- except no one uses these anymore and they are dead outside of legacy software.

The problem with the python tooling is no one can get it right. There aren't clear winners for a lot of the tooling.

13B is still super tiny model. Latent reasoning doesn't really appear until around 100B params. Its like how Noam reported GPT-5 finding errors on wikipedia. Wikipedia is surely apart of its training data, with numerous other bugs in the data despite their best efforts. That wasn't enough to fundamentally break it.

AI won't, but humans will to un-encumber AI

You don't want your agents to ask questions. You are thinking too short term. Its not ideal now, but agents that have to ask frequent questions are useless when it comes the vision of totally autonomous coding.

Humans ask questions of groups to fix our own personal short comings. It make no sense to try and master an internal system I rarely use, I should instead ask someone that maintains it. AI will not have this problem provided we create paths of observability for them. It doesn't take a lot of "effort" for them to completely digest an alien system they need to use.

This has been the fundamental issue with the 2.5 line of models. It seems to forget parts of its system prompts, not understand where its "located".

I assume its tool calling and structured output are way better, but this model isn't in Studio unless its being silently subbed in.

So many of these concepts only make sense under the assumption that AI will not get better and humans will continue to pour over code by hand.

They won't. In a year or two these will be articles that get linked back to similar to "Is the internet just a fad?" articles of the late 90s.

Grok 4 fast is a legit model. Their code models, including supernova still aren't smart enough. Claude and Codex are ahead. Its definitely fast, but who cares if you have to re-prompt it or it hits issues it can't fix.

I appreciate these takes, but I can't help to think this is just the weird interim time where nothing is quite good enough, but in a year an article like this would clearly be "overthinking" the problem.

Like when those in the know could clearly see the internet's path to consuming everything but it just hasn't happened yet so there were countless articles trying to decide if it was a fad or not, a waste of money, bad investment, etc.

GPT-OSS models are amazing, and a lot of the bad press was poor implementations of them in the usual tools and people not understanding how to handle their unique quant out of the box approach. Unlsoth has done amazing job unpacking best approaches

Flash-Lite is a seriously good model. I have had zero structured calls fail with it as its cranking out obscene tok/s. If you can run with something that isn't quite bleeding edge smart, this model is gold.

this made me laugh

I almost never have to reprompt GPT-5-high (now gpt-5-codex-high) where I would be reprompting claude code all the time. It feels like its faster, doing more, but its taking more of the developers time by getting things wrong.

Its been better for awhile, people are sleeping on it, just like they slept on claude code when it initially came out.

There are plenty of people who complain because its baked into their personality. Those people can also work at a bad place and have a target rich environment so to speak.

The only sure fire way to avoid this label being applied unjustly is always bring solutions, not complaints. Document your solutions and let it rest if leadership doesn't agree.

The male voices seem much worse than the female voices, borderline robotic. Every sample of their website starts with a female voice. They clearly are aware of the issue.

I think everyone is looking for back and white switches. Either coding agents are writing your code or they aren't. Humans will always be in the mix in some form, but the amount and skills they use is going to be radically different as time goes on.

I personally haven't written any significant code by hand since claude code landed. I also have a high tolerance for prompting and re-prompting. Some of my colleagues would get upset if it wasn't mostly one shotting issues and had a really low tolerance for it going off the rails.

Since gpt-5-high came out, I rarely have to re-prompt. Strong CI pipeline and well defined AGENTS.md goes an incredibly long way.