OpenAI just shook up the world. (Again)

Earlier today, OpenAI announced its first text-to-video feature. My initial take and reaction is "Amazing. Incredible. Absolutely stunning."

Look at the detail, granularity, and real-world simulation in these videos. For example, observe the reflection in a puddles in some of these videos and how the water moves. Achieving this level of detail is incredibly difficult for generative AI models at present, and these early takes are absolutely stunning. No one is close to this!

I am looking forward to reading the technical paper that is supposed to be out later today, and I will have more commentary for you today or tomorrow.

In the meantime, enjoy this supercut of some of my favorite clips from the announcement today. Wow!