In a fascinating 6-minute video, Google has put together the most significant interactions between Gemini – its new AI model, as well as the most powerful created so far by the Mountain View company – and a human being engaged in various activities. The clip shows how Gemini is able to “reason” about different inputs: text, images, video, audio and so on. Precisely its “native multimodal” nature would make it, according to Google, capable of reasoning better about complicated problems and concepts compared to other AIs in circulation, starting with ChatGpt, the most popular of the generative AIs developed by OpenAI.
by Pier Luigi Pisa
Read the article: Meet Gemini, Google’s most powerful AI: it understands physics and mathematics