Raw LLM Responses

Nah, that's where we were when this video released. AI is solidly cracking the decimal point here, maybe a couple thousand folks in the world who can compete with an absurd time-handicap in their favor... but most people in (professional settings, too) do not produce anywhere near as actionable and solid code as the best AIs right now. I'll take the comments about hype/doom, doesn't make it less of a reality. Anything you can easily interface with via language is basically done better by LLMs at this stage, and any tangible concern will be ironed out soon enough, as has always been the case so far. Which is exactly the reason why every coder on this planet is paying for at least a 20-USD or so subscription. It's stupid not to at least do boilerplate gen, which, by the way, everyone casually accepted as a "oooh, that's not really AI"-kinda thing to boot - but even the most skeptical folks, much like Prim, have long come around and at least are able to see the trajectory... if they really are still on the fence about current usability. Makes me wonder what people will circle back to griping about with AI. I mean, IP and licensing and all that? Ultimately moot when those models basically already are propping themselves up, what with synthetic data going from "very useful" to "sustainable as a monoculture", same for the already misrepresented issue of energy and water; LLMs were more efficient at comparable outputs a generation ago, now it's just ridiculous to even bring up. I'll take the entertainment stuff though. Things like AlphaGo, Robotlympics in various disciplines, man vs. machine. 10 40x coder legends competing against an agent who has to write compiling code in limericks.