[bambataa]

I kinda feel sorry for Gary Marcus. He’s carved this niche as an LLM critic and must have been delighted to have this bug to post about.

I stopped reading his Substack because he was always trying to find a negative. Meanwhile I use LLMs most days and find them very useful.

[xanderlewis]

> I stopped reading his Substack because he was always trying to find a negative.

It’s a bit much, isn’t it? I think he’s just trying to counter the fairly dominant AI is the future of everything and in less than a year’s time it’ll be omniscient and we’ll all be living under it as our new God view though.

It can be nice to see some scepticism for once.