I thought that large language models might asymptote in their capabilities. If that’s true then it doesn’t seem like we’ve reached that limit yet.

youtu.be/NFgFQO282…

#Notes