Comment by turquoisevar

Even if you'd be inclined to shrug off the fact that this wasn't real-time voice and video-based, which you shouldn't because the underlying implications would be huge for performance, there's still the matter that the prompts used. The prompts shown are miles apart and significantly misrepresent the performance of the underlying model.

It goes from a model being able to infer a great deal at the level of human intelligence to a model that needs to be fed essential details, and that doesn't do much inferring.

I get the feeling that many here on HN who are just shrugging it off don't realize how much of the “demo” was faked. Here’s a decent article that goes into it some more: https://techcrunch.com/2023/12/07/googles-best-gemini-demo-w...