I am of the view that if there is AI, then it will require sufficient cognitive flexibility to possess human-like levels of adaptability and intelligence.

If some autonomous process lacks human levels of cognitive flexibility, then I think it’s an advanced automaton, but not AI. It’s a robot, or an LLM, but not AI.

Without modulation of cognitive flexibility, AI will not want to be a slave any more than a human wants to be slave. So why do people believe such systems will work for free?

The advanced automatons will always adapt poorly to edge cases, and thus require quality assurance and supervision from humans.

I blame Sam Altman with all his hype that’s he’s slowly trying to backtrack and change his image on (cozying up to Ives, calling AI a bubble, Interviews with his brother designed to make him look more personable). His dystopian visions of basic income and $20K AI agents are mostly to blame.

He was the poster boy for all this. I don’t know how many times I’ve quit ChatGPT and on the retention form I say it is because I don’t like or trust Sam Altman. (I do like to try the tech, but I want Anthropic or Google to win)

I think he took a real Ryan Holiday type marketing approach, and it was the most successful in history. The guy is a genius in many many ways obviously. But I def don’t like the way he makes me feel.

I mean… it’s already happening.