If you look at nvidia’s presentation, the next “big thing” is robotics and related software. Well, it’s kind of obvious, you can only do so much with just software (AI) even if it works as best as it can, it is bound to the digital world.
If getting LLM’s (alignment, accuracy etc.) right was challenging, then the same on top of dynamic 3D environments will result in an even bigger shit show.
My issues are fundsmentally two fold with gen AI:
Who owns and controls it (billionares and entrenched corporations)
How it is shoehorned into everything (decision making processes, human-to-human communication, my coffee machine)
I cannot wait until finally the check is due and the AI bubble pops; folding this digital snake oil sellers’ house of cards.