Google demos new artificial intelligence model Gemini, which leaves many of us speechless It turned out that he had cheated by deliberately deceiving us..
The artificial intelligence (AI) race has begun, and OpenAI has knocked large companies out of their game. Apple continues to do so without any haste. But it seems Creating my own model which we can start to see with the next iPhone 16 model, but It looks like Google has rushed through and showcased what Gemini can do.His new AI model with video is impressive, or rather if it’s true he cheated.
In a video that has received nearly two million views in less than 48 hours, we get to see how Gemini works in a way we’ve never seen before until now. Using AI live video and audio Google answers questions with our voices. By knowing at all times what is happening in the pictureAt the same time, it happens without waiting long. If you spend just a few minutes watching the video you will no doubt be impressed.
But it turns out that’s not how Gemini works. Google’s models don’t recognize the video footage we see, but they do. What is really perceived are still images and written questions.It has nothing to do with what is shown in the video, for example. To give you a good idea of what I mean, at 4:27 we see how three paintings are presented (Sun, Saturn, and Earth) and are asked. whether the order is correct or not Gemini responded that this was incorrect and gave additional royal orders. But that’s not what happened. What Gemini sees is a photograph of three stars and a written question: “Is this the correct sequence? Consider your distance from the sun and explain your reasoning.
The entire Google video is a montage that has nothing to do with how Gemini actually works. It’s true that Google’s AI gives us the answers we see in the video and applies the images we see in the video, but The “how” is very important and here Google has deceived us.. Showing us how this model works is the right thing to do. But it wouldn’t be very impressive, would it?