why I think the "chatgpt era" of AI is already hitting a wall

ngl, the obsession with just making LLMs bigger and hoping they stop lying to us is getting old. it feels like we’ve reached the limit of what "fancy autocomplete" can actually do for society. like, u cant run a power grid or design a microprocessor on a model that might decide to hallucinate just because the prompt was worded weirdly I was checking out the speaker list and panel notes for the Milken Conference and it’s pretty telling who they’ve got on stage this year. seeing the ASML and Google guys sit down with Logical Intelligence to talk about "deterministic" AI makes it feel like the pivot is finally happening in the background the future isn't just a smarter chatbot. it's gonna be about these energy-based models that actually understand constraints and mathematical logic. The industry is finally moving from "AI for fun" to "AI for stuff that literally cannot fail" bit of a reality check for the silicon valley hype cycle but honestly, it’s a relief to see some focus on correctness for once