yeah, they really sound like they're all high on their own supply.
however if they've really got something that can eventually solve math problems better than wolfram alpha / mathematica that's great, i got real disappointed early in chatgpt being entirely useless at math.
lemme know when the "AGI" gets bored and starts grinding through the "List of unsolved problems in mathematics" on its own and publishing original research that stands up to scrutiny.
THIS. If it has the whole corpus of research, math, physics, science etc to know how the world works and know it better than any human alive, it should be able to start coming up with new theories and research that combines those ideas. Until then, it's just regurgitating old news.
however if they've really got something that can eventually solve math problems better than wolfram alpha / mathematica that's great, i got real disappointed early in chatgpt being entirely useless at math.
lemme know when the "AGI" gets bored and starts grinding through the "List of unsolved problems in mathematics" on its own and publishing original research that stands up to scrutiny.