"AI is overhyped, and hasn’t had much real world impact. That’s what Samuel James argued in a recent Substack post"
My post:
"Look, I am not saying, and I don’t believe, that AI is a nothingburger. It’s disruptive, yes. It will change industries, yes."
Once again, I only ask AI advocates to represent opinions like mine accurately. But so many refuse. Nobody can blame me if I start asking why that might be.
Apologies Samuel, it's not my intention to misrepresent what you said, and I welcome the pushback.
You did also say "AI’s potential to transform society as a whole has been talked about nonstop for at least three years. And so far, the main thing we have to show for that are porn bots and data centers."
Given the examples I cite of the capabilities of LLMs, I think this is incorrect, missing current applications of AI e.g. to write code or help solve physics problems. (I know you have a caveat about this in your post, but what evidence would you accept of AI's impact given that you don't use the tools?)
I also tried to stress that greater AI capabilities are not merely a matter of guesswork but predictable based on scaling laws https://en.wikipedia.org/wiki/Neural_scaling_law (which I don't necessarily think will hold forever).
"AI is overhyped, and hasn’t had much real world impact. That’s what Samuel James argued in a recent Substack post"
My post:
"Look, I am not saying, and I don’t believe, that AI is a nothingburger. It’s disruptive, yes. It will change industries, yes."
Once again, I only ask AI advocates to represent opinions like mine accurately. But so many refuse. Nobody can blame me if I start asking why that might be.
Apologies Samuel, it's not my intention to misrepresent what you said, and I welcome the pushback.
You did also say "AI’s potential to transform society as a whole has been talked about nonstop for at least three years. And so far, the main thing we have to show for that are porn bots and data centers."
Given the examples I cite of the capabilities of LLMs, I think this is incorrect, missing current applications of AI e.g. to write code or help solve physics problems. (I know you have a caveat about this in your post, but what evidence would you accept of AI's impact given that you don't use the tools?)
I also tried to stress that greater AI capabilities are not merely a matter of guesswork but predictable based on scaling laws https://en.wikipedia.org/wiki/Neural_scaling_law (which I don't necessarily think will hold forever).