Our most expressive and steerable TTS model yet! Designed to give builders granular control over AI-generated speech, Gemini 3.1 Flash TTS is really fun to play with! Available in preview today - for devs via the Gemini API & @GoogleAIStudio + for enterprises on Vertex AI
Simulations are the future, & one of the main tools we’ll ultimately use to understand and predict things about the universe. This is why I’m so excited about Genie 3, our latest interactive world simulator - here are some insanely cool things you might have missed about it 🧵:
We use Genie 3 to create worlds to train other agents in! We give our SIMA agent goals to achieve in environments generated by Genie 3, & watch as Genie 3 responds to the actions that SIMA takes. Basically it's an AI playing in the mind of another AI! 🤯
Genie 3 shows ‘inception-like’ capabilities. You can prompt it with a video & text, and it will try to keep things consistent. Here @jkbr_ai uploads a video of himself using Genie 3, while asking for a T-Rex in a jungle setting, & the results were amazing. Notice how first Genie modifies the display on the TV & laptop screens to show the T-Rex, and then when we look out of the window it has transformed the outside into the jungle! x.com/jkbr_ai/status…
Official results are in - Gemini achieved gold-medal level in the International Mathematical Olympiad! 🏆 An advanced version was able to solve 5 out of 6 problems. Incredible progress - huge congrats to @lmthang and the team! deepmind.google/discover/blog/…
We achieved this year’s impressive result using an advanced version of Gemini Deep Think (an enhanced reasoning mode for complex problems). Our model operated end-to-end in natural language, producing rigorous mathematical proofs directly from the official problem descriptions – all within the 4.5-hour competition time limit! We'll be making a version of this Deep Think model available to a set of trusted testers, including mathematicians, before rolling it out to Google AI Ultra subscribers.
Btw as an aside, we didn’t announce on Friday because we respected the IMO Board's original request that all AI labs share their results only after the official results had been verified by independent experts & the students had rightly received the acclamation they deserved
Excited to share #AlphaMissense our new AI system that can classify whether genetic mutations (missense variants) are benign or harmful - a critical step toward uncovering causes of many diseases, from cystic fibrosis to cancer. In @ScienceMagazine today dpmd.ai/AlphaMissenseDH
Just as swapping a single letter can alter an entire word, in the language of DNA, swapping a single amino acid can alter an entire protein. These swaps are called missense variants. An average person has over 9,000 missense variants - most benign, but even one can cause disease
Our Missense model, building on #AlphaFold, has helped classify 89% of all 71 million possible missense variants to predict if they’re benign or pathogenic. Until recently human experts had classified just 0.1% of them