Seeing a lot of posts about Kimi K3.1, especially in the game dev world. After reviewing some this is nothing more than a scripted marketing push by influencers. The 3d models are actually Astra, labeled and compared against Claude (Claude is notably worse than OpenAi at 3d).
@ZviGoldstein@CrazyShyyt Every human brain is capable of prginal thought and drawing conclusions without being trained on that particular conclusion.
Ex. Religion.
@Dabeer_builds@ZviGoldstein@CrazyShyyt You could've written a c program 40 years ago to solve them, by that I mean runs millions of simulations until you have a match. Is that AGI?
@Dabeer_builds@ZviGoldstein@CrazyShyyt I guess we're all smarter than AGI if we toss turds at the wall and one sticks, right?
14 problems out of millions of attempts passed. They claimed 300 were solved. 4 people were happy with the results but agreed the algorithms were incomplete to be accepted as solved.