Can see Wolfram's approach here - interesting how he's coming to generative AI with a beginner's mind, from first principles. He makes a lot of interesting poin

Can see Wolfram's approach here - interesting how he's coming to generative AI with a beginner's mind, from first principles. He makes a lot of interesting points, but I can't help and think that he would save a lot of time talking to someone who knows the field better, instead of figuring it all out on his own. The stuff about attractors and dimensions in model space is missing the architectural understanding (layers representing features, backprop)