Just used a text to video AI model to generate a cartoon dancer! Quality isn't perfect yet, but it's exciting to see the potential of it. Can't wait to see how much it improves by the end of the year! #modelscope #text2video #AIArt

Any #AI model intended for human #interpretation (#consumption) has a #TheoryOfMind baked into it. However until #Cicero by #Meta models were static and rather primitive, and mostly implicit.

The #TheoryOfMind that is in every model is the one which designers had in their minds - implicitly or explicitly - when designing and testing their model.

As the adage goes: humans cannot not communicate.

#LLM #Text2Image #Text2Video #Text23D

Hi all, I'm Ruben Villegas and I'm a researcher at Google Brain. I work on generative models and self-supervised learning in the video domain. I'm one of the authors of Phenaki (phenaki.research.google) recently highlighted in the Google AI event 22' (https://youtu.be/X5iLF-cszu0). #Phenaki #Text2video #Video

Happy to join everyone here and learn about and discuss all the awesome research being done by the community