It also aligns with what I saw from Claude & Claude Code when I used it last year for a while. Now I use Codex and don't see (nearly) as much of that sort of behavior.

> But LLMs seem to have a hard time getting the big picture and reusing code that is already implemented and ALMOST does what you want vs. rewriting everything from scratch.

Yeah that's probably the weakest point of LLMs still. GPT-5.6 Sol got much better at this for me. I still usually end up doing 2-3 iterations of prompts and exploration to clean up various things but usually it's pretty light work now.

> hundred of thousands of lines of documentation written as walls of text in markdown and weird coding decisions, your codebase is impossible to work out for a human.

Okay Claude definitely seems to have an issue there. From what I've seen from coworkers using Claude it generates reams of endless docs. I resist the urge to `rm docs/planning*.md`! I don't think they get how bad Claude is at that.

A month back I tried the latest DeekSeek and it made reams of text back and forth with itself, but the code it output was reasonable and it didn't make pages of markdown files either.

> I make it build things in small changes, I give very specific instructions about the architecture, and I make sure to point out existing functionality that can be used instead of writing something from scratch.

That's a bummer. I found that once the models start having problems that it cascades.

I've also been able to keep steady progress on a 70k+ LOC GUI side project without the endless whack-a-mole of bugs using Codex and Sol. Squash a bug, review architecture, move on, etc.

That’s what I got from reading about a third of that exchange.

I mean that’s clear from reading anything about the dude, but to _experience_ it so to speak is different.

A true mark of brilliance is someone who makes something difficult seem easy by framing it just right.

In this instance ChaGPT is just expanding upon his prompts. Not to diminish, it’s amazing what it’s capable of.

The parent wasn't referring to individual LLMs developing, but that overall new SOTA LLMs models are increasing in capabilities in ways similar to how human children do.

These SOTA LLMs aren't trying to mimic existing children's drawings, but interestingly they're following somewhat similar progression that human children do as they develop.

For example I had Codex port Kiwi Cassowary C++ to Nim: https://github.com/elcritch/kiwiberry

Might have to try that myself. Do you still use your phone/tablet/etc or does it also encourage you to use them less?

> Also have the same experience fighting depression-like symptoms and anxiety. It just takes a lot of time and is difficult. Some people just don't have these problems and I do but this doesn't mean I am genetically attuned to be like this and I can't do anything. It is just difficult.

Should I mention that "neurodivergence" and different sleep pattern genes have a large co-morbidity? E.g. many people with anxiety / ADHD / dyslexia / depression / etc have a very high likelihood of having delayed sleep or other genes.

Last year I did an experiment of sorts while unemployed for a time and found that if I just slept and woke when tired that my sleep time would naturally recess and eventually "flip" after about a month.

My entire life I've wondered why I feel incredibly tired and found waking up so difficult. Turns out that if you follow your bodies dominant sleep cycle it's a synch to wake up. Unfortunately, it doesn't work with modern life very well.

It's on the same topic, possibly spurred by the CC. Yes it uses the same graph (which I presume CC also didn't make), but otherwise the content and writing is very different. It's less detailed, possibly LLM written, but also more approachable.

I guess re-posting about a salient topic is now just "AI copyright theft".

From what I've read, the little plastic baubles part of buttons are cheap, it's the electrical wiring harnesses and assembly that are expensive part of physical controls.

Still a scam can include other "deceptive schemes" – maybe he's just trying to get fame or plain old attention.