This looks interesting, but can someone explain to me how this is different from video generators using the previous frames as inputs to expand on the next frame?
See the demo on their homepage. Calling it a world simulator is a marketing gimmick. It's a worse video generator but you can interact with it in real time and direct the video a little bit. Next version of this thing will be worth looking, this one isnt.
There is soo much marketing bs around these things it drives me nuts. and it doesn't help that the large labs and credible individuals like denis use these terms. "world models" are video generator with contextual memory but that term is soo misplaced. when one thinks of a "world model" you expect the thing to be at least be physics engine driven from its foundation, not the other way around where everything is generated and assumed at best.
Based it on other video models, all the ones I have seen keep improving. This one should too. Infact, Google is doing it already with their Genie (IIRC). That one is high quality and interactive.
Is this more than recursive video? If so, how?