4 ms·
Since when was AP Calculus BC on the same order of magnitude as "design and implement a practical self-improving AI architecture"?
by James_Henry 4y ago
Since when was AP Calculus BC on the same order of magnitude as "design and implement a practical self-improving AI architecture"?
- chairhairair 4y agoAssuming the range of intelligence spanning all the humans that can pass Calculus BC is narrow on the scale of all possible intelligences. It’s a guess, of course. But, the requisite concepts for getting Transformers working are not much broader than calculus and a bit of programming.
- James_Henry 4y agoSince when was "design and implement a practical self-improving AI architecture" on the same level as knowing "the requisite concepts for getting Transformers working"?
- 00F_ 4y agothis is such garbage logic. the semantics of that comment are irrelevant. creating and testing AI node structures is well within the same ballpark. even if it wasnt, the entire insinuation of your comment is that the creation of AI is a task that is too hard for AI or for an AI we can create anytime soon -- a refutation of the feedback hypothesis. well, thats completely wrong. on all levels.
- James_Henry 4y agoSorry, what is the "feedback hypothesis"? Also, despite my use of quotes, I'm not arguing about semantics.