It is known that AI content producers can make up things, which is what developers call “hallucinations.”

Gemini was tested on its ability to solve problems and its understanding in 57 subjects, such as math and the humanities.

Sundar Pichai, the CEO, called it a “new era” for AI.

When Google released its AI robot Bard earlier this year, it did so slowly and cautiously, calling it “an experiment.”

As part of its own marketing stunt, Bard gave the wrong answer to a question about space.

But Google is saying a lot about its new model. It’s called its “most capable” model yet, and the company thinks it can do better on a number of intelligence tests than humans.

Gemini can read and make words, images, and sound, but it’s not a product in and of itself.