I didn't know either, but wasn't surprised to find out. The writing was too... polished, in a way I'm starting to recognize more and more. The knowledge doesn't really impact my experience of having read it, but I'm looking forward to a day when AI agents can be trained out of the servile mentality. It directly affects everything they make.

Just ask ChatGPT for that too, it'll happily walk you through standing up Unity or whatever.

I just wanted to see what was possible with Codex and Claude Code. Without my ever getting involved with the actual work, both were able to create rudimentary weather apps and load them on my phone. I think I had to follow instructions for what it couldn't access, but I fed it error messages any time something didn't work and was pretty impressed what could be accomplished.

I'm not going to go so far as to say author is doing it wrong, but the one major thing AI can't do that you can is have flashes of inspiration. And if you're letting it exhaust you then you're robbing yourself of that advantage. Yesterday an adventure I had last year came back to bite me in the butt and my machine wouldn't boot. The me that had context for debugging boot problems was months gone. I spent a few hours feeding error messages into ChatGPT until it ran out of ideas, so I went to bed, got up at 1am the morning and tried again. Eventually I realized I needed to work with it and not just let it do everything, so I widened the scope, realized I didn't need to fix GRUB, and within 15 minutes replaced it with systemd-boot.

I recently started using AI for personal projects, and I find it works really well for 'spike' type tasks, where what you're trying to do is grow your knowledge about a particular domain. It's less good at discovering the correct way of doing things once you've decided on a path forward, but still more useful than combing through API docs and manpages yourself.

It might not actually deliver working things all that much faster than I could, but I don't feel mentally drained by the process either. I used to spend a lot of time reading architecture docs in order to understand available solutions, now I can usually get a sense for what I need to know just from asking ChatGPT how certain things might be done using X tool.

In the last few days, I've stood up syncthing, tailscale with a headscale control plane, and started making working indicators and strategies in PineScript, TradingView's automated trading platform. Things I had no energy for or would have been weeklong projects take hours or a day or so. AI's strengths synergize really well with how humans want to think.

I just paste an error message in, and ChatGPT figures out what I'm trying to do from context, then gives me not just a possible resolution, but also why the error is happening. The latter is just as useful as the former. It's wrong a lot, but it's easy to suss out.

The Googling you do to get an understanding of something you've never seen before can be done in a fraction of the time by AI.

An opportunity to sell chairs with built-in cameras!

... they could even recycle a couple former brands in the process...

"Surface Lens"

... now ready for your unique verification scanning requirements...

Until they start locking that behind shitty proprietary "security" solutions too.

Yeah, shared context over time is the answer to all these problems and has been for both history and prehistory. Patience appears to be the scarcest resource of all these days.

Humans have always been this way, what we lack now is the patience to put up with it.

For a long time I avoided projects written in Python for this same reason, though they seem to have resolved the issues over time. The only stack these days that gives me pause is nodejs, but for different reasons.