mondobe

LinkedIn has seriously deranked posts with links in the past. Everything within Meta is heavily A/B tested, but it seems feasible that this will become real.

This absolutely puts a lot of what I've been feeling into concrete, scientific terms, thanks!

I actually thought of this song while writing the post. I was definitely feeling my hubris at that moment, but I would at least liked to have given it a try in 2020-2022, when WebAssembly/Rust on the web was getting good and we were all stuck inside.

Thanks! I have more substantive thoughts about AI I'd like to write down later.

Very insightful and encouraging words, thank you!

I have no problem with creative destruction, elevator operators died out for a reason. If there was a group of people who rejected automatic elevators for some reason, it stands to reason there would be a company to snatch up that market. (case in point: no-touch Shabbat elevators for Orthodox Jews)

Good point, and one I've thought about a lot. Right now, I think we're at a stage in the hype cycle where, for the overwhelming majority of founders, appealing to AI hype is much easier than lying to appeal to anti-AI sentiment. If and when this changes, it might manifest as an incentive for companies to be transparent with their source code (either fully open-source or at least "source available" or auditable).

Thanks for the perspective! I tend to consult ChatGPT with high-level queries as a replacement for Google, and I get a lot of mileage out of combining that with plenty of good old documentation-reading.

Fair point about game development. I'm definitely looking more towards creative tools (like music production apps) because making a game engine is just a Herculean task. I've found that velocity is good (i.e. one meaningful feature every day or two) in the early stages of building for the web without AI.

I'll trust you on this subject, anal_reactor