thrwjud

The more compute you need to get state of the art performance the closer to the plateau you are. If you didn't need the compute researchers would be getting better results with smarter training. Given that the gpt family of models need more energy to train than Nevada needs to keep the lights on they are very much on the flat part of the logistic growth curve.

Every deep learning tech had an exponential growth phase, followed by a slowdown followed by a platou the nothing could break until a fundamentally new architecture came along. People get excited about the first part, project it into the second and start companies by the time we're well into the third.

>Nuance and subtlety is impossible to distinguish from dog whistling. Most users don't understand that by "dunking" on someone, they are actually promoting the content. In summary, it gave me a darker feeling about the world in general, and in exchange, did not offer much to enrich my life. [...] And finally, I have more nuanced conversations with my wife and friends about politics and other hot topics, without worrying that I'm going to come off as dog whistling or naysaying.

It sounds like the problem with social media for OP is between the ears rather than on the screen.