Ah, the long form article - these should be published in parts, maybe this one in four.

The Dunbar number is 150 for humans but that only measures maintaining a group, maybe the behave nicely number is smaller.

Even though I use LLMs for code, I just can't read LLM written text, I kind of hate the style, it reminds me too much of LinkedIn.

I had to stop reading about half way, it's written in that breathless linkedin/ai generated style.

Interesting. Can AMD GPUs do direct io like this?

Only use this stuff in devcontainers, I find it mad people give this stuff this sort of access.

(I only use devcontainers for this purpose, I'm not really a fan in general)/

"Writing code is a solved problem" disagree.

Yes, there are common parts to everything we do, at the same time - I've been doing this for 25 years and most of the projects have some new part to them.

Novel problems are usually a composite of simpler and/or older problems that have been solved before. Decomposition means you can rip most novel problems apart and solve the chunks. LLMs do just fine with that.

not AI, but IA: Intelligence Augmentation.

AI is always bad at trains, I'm sure if the picture was wider there would be no gangway.

Oof that was a very unwise blog post to make.

Copywriter aside it looks like an interesting blog post.