🔍 BIG GOOGLE NEWS. GEMINI.

Google has introduced its multimodal AI, which can reason and better understand human intentions. It can work with videos, write text, create images, code, and music. Communication with the bot goes beyond chats and interfaces. It's a serious competitor to ChatGPT, outperforming it in several multimodal tests. We attach an amazing Gemini video presentation.

Gemini is the first AI that has beaten humans in a big language test called MMLU (Massive Multitask Language Understanding), scoring 90%. It's getting closer to AGI.

❓ What can it do?
This AI can handle different types of content: texts, images, audio, and video. Google even trained it on YouTube videos.

You can now use the Gemini Pro version for free (the video shows the Ultra model, which is more advanced but not out yet). Google plans to charge for Gemini with subscriptions in the future.

▶️ How to use Gemini Pro outside the US? Just follow these 3 steps:
1. Change your Google account language to English (United States) here.
2. You must use an American VPN. You can find an extension for this in Chrome.
3. Go to the Bard chatbot. You'll see a message at the top saying, "Bard has been updated in English with Gemini Pro."

Gemini comes in three versions:
– Nano for simple tasks, already available on Pixel 8 Pro smartphones;
– Pro for more complex tasks, available in the Bard chatbot from today;
– Ultra, the most advanced model, beats GPT-4 in tests and will be released next year.

Developers can access the Gemini Pro API starting December 13th. Coming soon on @GPT4Telegrambot 🔜

The most potent version, Gemini Ultra, will be in Bard and available via cloud API in 2024.

Interesting fact: Gemini is named after the merging of Google's two main AI labs and refers to NASA's Gemini project, which helped start the moon landings in the Apollo program.

#Google @hiaimediaen