What I mean by “nowhere near human intellects” is that e.g. Claude is exclusively good at coding. Ask Claude to paint, or walk (using a robot body), or understand body language, and you’ll see how limited and “idiot savant” Claude is.
Also, Claude understands patterns within the code, not the actual code itself. I suppose the same is true for non-coders as well, though.
But as I stated before, Claude does use a neural network. With enough time, effort and incentives, an AI like Claude could become general purpose. And at that point, we’d be approaching true sentience and sapience.
What I mean by “nowhere near human intellects” is that e.g. Claude is exclusively good at coding. Ask Claude to paint, or walk (using a robot body), or understand body language, and you’ll see how limited and “idiot savant” Claude is.
Also, Claude understands patterns within the code, not the actual code itself. I suppose the same is true for non-coders as well, though.
But as I stated before, Claude does use a neural network. With enough time, effort and incentives, an AI like Claude could become general purpose. And at that point, we’d be approaching true sentience and sapience.