Muslims are expected to pay taxes, follow rules, and vote—just like everyone else. But when democracy actually works and one of them wins, people lose it. I don’t believe in religion, but I do respect all. @ZohranKMamdani

The most beautiful democracy in the world. 🇺🇸

Repeat after me: 🇺🇸 The USA has no official language. 🇺🇸 The USA has no official religion. 🇺🇸 The USA is built by immigrants. That’s what makes this country home.

The Illustrated NeurIPS 2025: A Visual Map of the AI Frontier New blog post! NeurIPS 2025 papers are out—and it’s a lot to take in. This visualization lets you explore the entire research landscape interactively, with clusters, summaries, and @cohere LLM-generated explanations

I really like this guy! hope you give him your vote!

(1) Introducing the AI for Math Initiative! Supported by @GoogleDeepMind and @GoogleOrg, five leading global institutions (@imperialcollege, @the_IAS, @Institut_IHES, @SimonsInstitute and @TIFRScience) are coming together to pioneer the use of AI in mathematical research.

“A Couple of Years Away” itself is a couple of years away!!!

We're at a point where basic kindness is so rare, it's met with shock or suspicion. Wild!

We think you're going to like these updates a bunch 🍌 Introducing six new visual styles for Video Overviews (powered by Nano Banana) and a shorter "Brief" option for quick insights. These will start rolling out to Pro users this week and to all users in the upcoming weeks.

Are we overfitting to MatMul-centric hardware? My new post with Jan Wassenberg explores how this risks creating an evolutionary dead-end for AI hardware, potentially trapping us in an innovation cul-de-sac. Let us know what you think! https://t.co/JHxhUr16So

I love this visual so much

This is what happens when you just teach how to prompt AIs. We also need to upskill people on how to assess, evaluate, or **verify** AI outputs. Domain expertise matters for this. As @balajis best puts it, "AI verifying is as important as AI prompting."

When you build in silence, people don’t know what to attack.

From the beginning, we had clear goals for Chrome: simplicity, speed, and security.  And now with AI, we have a chance to make the browser even safer, more intelligent and helpful.  Our next Chrome era is just beginning…

Incredible milestone: an advanced version of Gemini 2.5 Deep Think achieved gold-medal performance at the ICPC World Finals, a top global programming competition, solving an impressive 10/12 problems. Such a profound leap in abstract problem-solving - congrats to @googledeepmind!

Q: How do you define the meaning of a program? A: With semantics! How program output is computed from input, and checking of rules. Watch my new Programming Language Pragmatics video, covering dynamic vs. static semantics, & how to represent a program with abstract syntax trees.

"Learning to break down problems into smaller pieces is one of the most important skills in computer science/life." — Addy Osmani (@addyosmani)

Models converge. Apps defend the moat.

Our image editing model is now rolling out in @Geminiapp - and yes, it’s 🍌🍌. Top of @lmarena’s image edit leaderboard, it’s especially good at maintaining likeness across different contexts. Check out a few of my dog Jeffree in honor of International Dog Day - though don’t let

Image generation with Gemini just got a bananas upgrade and is the new state-of-the-art image generation and editing model. 🤯 From photorealistic masterpieces to mind-bending fantasy worlds, you can now natively produce, edit and refine visuals with new levels of reasoning,