Emanuel’s review of Do People Understand Anything? A Counterpoint to the Usual AI Critique > Likes and Comments
2 likes · Like
Manny wrote: "I also like the submarine line and use it often :)
Our basic point in the paper is simple: we invite readers to consider just what they believe people are doing when they "understand", and why AIs..."
I think you're right (if I've understood you correctly), which probably makes me the wrong audience for this argument. At this stage, the burden of proof lies with those who claim inherent boundaries to AI capabilities. We've repeatedly seen claims of the form 'AI will never be able to do X' die with model scaling. Any compelling argument for limits needs to go beyond simply asserting that human cognition ≠ machine cognition or invoking 'true understanding.' Nonetheless, keep up the good fight :)
back to top
date
newest »
newest »
Manny wrote: "I also like the submarine line and use it often :)Our basic point in the paper is simple: we invite readers to consider just what they believe people are doing when they "understand", and why AIs..."
I think you're right (if I've understood you correctly), which probably makes me the wrong audience for this argument. At this stage, the burden of proof lies with those who claim inherent boundaries to AI capabilities. We've repeatedly seen claims of the form 'AI will never be able to do X' die with model scaling. Any compelling argument for limits needs to go beyond simply asserting that human cognition ≠ machine cognition or invoking 'true understanding.' Nonetheless, keep up the good fight :)
Our basic point in the paper is simple: we invite readers to consider just what they believe people are doing when they "understand", and why AIs can't do the same thing. My impression is that many critics of AI are reluctant even to consider these questions.