The technological struggles are in some ways beside the point. The financial bet on artificial general intelligence is so big that failure could cause a depression.
The technological struggles are in some ways beside the point. The financial bet on artificial general intelligence is so big that failure could cause a depression.
i’ve been saying for months that every billion we spend on LLMs etc. is moving us further away from AGI, not closer, and that retreat is accelerating.
it feels so wildly irresponsible for evangelists like ellison to continue to hype this dumpster fire up. but then he doesn’t stand to lose his mansions, his islands, his jets.
I remember years ago before all this LLM hype went super saiyan talking to a guy who was studying to go into artificial intelligence (this was at some fairly prestigious northeast university). He clarified after some conversation that no he wasn’t interested in machine learning and neural net style artificial intelligence, he was interested in the extension of logic based systems into more and more sophisticated means and of what kinds of rules and architecture best facilitated that.
I remember thinking something along the lines of “Trying to make intelligence with rules! That seems outdated!” at the time because I didn’t know much about the topic other than that machine learning seemed to be drawing a huge amount of excitement.
I think about that moment sometimes and ponder how much of a fool I was, that guy was totally right, and you are too, we are running full speed away from most meaningful avenues of pursuit towards creating sentient artificial intelligence AND creating automated systems that can actually be entrusted with serious responsibility not to violate basic priniciples or hallucinate bullshit and hurt someone in the process… or just break constantly.