I dabbled a bit in ML before GPT, and when the most recent hype-rocket launched I did a deep dive into LLMs, and I gotta say…

None of my hopes or horrors regarding “AI” have changed much along the way.

It’s pretty much the same thing we’ve been doing since the industrial revolution, which is to try to map human behavior onto mechanical processes so that we can optimize for from a quantitative, objective frame of reference.

GenAI is only unique in that it’s an especially mask-off moment for the ruling technocrats. We are destined to become wetware plugins for a capitalist machine whose goal isn’t even as interesting as turning everything into paperclips. It’s worse than a rogue superintelligence.