[haltist]

Can LLMs do this?

[vore]

I wouldn't use an LLM for anything that can be done 100% precisely, like this.

[haltist]

OK, just curious how LLMs are stacking up in logical tasks like this. I kept hearing we were close to AGI so just wondering how far there is to go.

[clbrmbr]

Humans can do these intersections, but we don’t do it by riffing off the top of our heads. We carefully develop and apply a formal system. LLMs are just (a very important) component.

[Izkata]

We've been "close" to AGI for like 40+ years.