In the past, I would’ve said no. I think the answer is more nuanced now. I think the current form of LLMs is somewhat AI in that, there isn’t really intelligence, but it was artificially built to be statically intelligent. AI in its current form with a static LLM black box isn’t intelligence in the way people are intelligent and constantly learning. In fact, it doesn’t really learn like a human, function like a human brain, etc. But it is “agentic” in that it can complete tasks.

Unless there’s a breakthrough in how LLMs function, not just learn, to be like the human brain, I don’t think we’ll achieve AGI. It’s possible the learnings of the LLM frenzy era will help blueprint what AGI will look like. I believe LLMs are like the speech center rather than intelligence. LLMs can formulate the right words for communication based on some signals. But the learning function isn’t there yet.