Google drops its trump card! The Genie 4 world model can hold real-time conversations—but don’t get carried away just yet.

On October 27, Google DeepMind officially unveiled its most powerful world model yet, Genie 4. This upgrade is a big leap forward:

For the first time, it natively generates audio and supports real-time conversations, with characters improvising responses based on your actions;

You can use natural language while it’s running to change the weather or switch camera angles in real time;

Physics simulation has improved significantly, and the duration of a single interaction has doubled to 2 minutes.
Gemini acts as the “director,” responsible for planning.

But note: This is currently only a research preview, available only to invited testers and, later, AI Ultra subscribers—not the general public.


A trader’s perspective: The tech giants’ race to build AI world models is fueled by an enormous appetite for computing power and energy. With macro liquidity already tight, this level of technological advancement will further reinforce the narrative of “AI sucking up all the resources.” Watch how tech stocks move tonight and how they correlate with the AI sector.

Disclaimer: This is only a summary of public information and an analysis of the tech narrative; it does not constitute investment advice. Crypto assets are highly volatile. Please make your own decisions.
#谷歌 $GOOGLB