If the problem with limited context will be fixed, then the first thing that will happen is that humans will become (limited) sensory inputs and action actors for the true AI; thus forming a biotech collective with artificial brain on top of the hierarchy.

Or if we view intelligence in the traditional sense of "being able to come to the Truth", then without experiencing the world, it will not be possible; random internet articles from a training set will not do.

And by "experience" I mean not only being plugged to some inputs but also to be able to react and change its own environment.

а що не так з сабстаком? через андерсена?

addyo.substack.com/p/vibe-codin...

хороша стаття на цю тему і коли ллм взагалі буває корисною

artificial - так. general - можливо, якщо не фантазує про те чого не знає. (human-level) intelligence - ні, тому що не може адаптуватись (як ми), не можу уявляти те чого не існує (як ми), не може бути самостійною (як ми).
це все ще статистична модель, машина з патернами без власної інтуїції

coz games are 80+$ ( ͡° ͜ʖ ͡°)

no way this will happen, you can overthrow the government only only on weekends, otherwise it's illegal!!!

Automating and outsourcing boring tasks sounds like very compelling idea. It's counterintuitive that people actually need boring stuff to live good

still prefer classical "sunset 1.0"...

yo. So where are the projects? 🤓