Why are you talking about how it feels once you’ve seen AGI when you’ve never seen AGI, Sam?

In all seriousness, we’ve got glorified autocorrect right now. Even suggesting any of these LLMs is actual AGI is laughable. I’m not saying they can’t do some interesting things, but unless Sam has access to models that are equivalent to what would be GPT-50 he should avoid throwing in buzzword acronyms for no reason.