Well, the idea here is that LLMs should progress towards AGI and thus could just solve any problem while a purpose-built chess AI has a very narrow scope of application. Of course it's clear for everyone they aren't anything like AGI currently.

Comments