AlphaGo won with the help of “Move 37,” a play so unconventional experts thought it was a mistake. That win helped herald the modern era of AI and led to a decade of scientific discovery.
Read the full reflection on the journey from AlphaGo to AGI by Demis Hassabis → goo.gle/4rjVjAH
10 years ago, Google DeepMind’s AlphaGo became the first program to beat a world champion at Go — a game with more moves than atoms in the universe. ⚫⚪✨
We launched it as a free, open-source tool a year ago, and today, research groups are using it to make sense of their camera data faster than ever. See how our open-source Earth AI model SpeciesNet is helping to promote wildlife monitoring and conservation worldwide. goo.gle/4riXtRk
A collage shows wildlife in North American forests. Each animal, including three black bears and a coyote, is framed by a colored box with a species label and a high confidence percentage.
The four photos show African wildlife in their natural habitat at various times of day. Digital overlays identify the animals, such as an "elephant: 99.8%" and a "lion: 99.4%," for data collection purposes.
SpeciesNet is an Earth AI model trained to automatically identify nearly 2,500 categories of mammals, birds and reptiles from data captured by motion-triggered wildlife cameras.
Google I/O, our annual developer conference, is back May 19-20. 🚀 Tune in to learn about our latest breakthroughs and updates in products across the company, from Gemini to Android and more with livestreamed keynotes and sessions at io.google.
Image with a black background showing a trophy on the right with text reading "Google I/O Make Build Unlock".
The #GoogleIO puzzle is officially live. 🧩
You can play, remix, and help the community crack the code in real-time.
Are you ready to help us reach the goal? 🤝
Start here → goo.gle/iopuzzle26-6
Thanks for flagging. This was a bug affecting book search results, and it's fixed now. Things should be back to normal.
4/ See how we're decoding life on Earth to protect it for future generations → goo.gle/4qckKn8
A collage of four vertical panels featuring different endangered animals in their habitats. From left to right: Panel 1: A cotton-top tamarin with a distinctive shock of white hair on its head, perched on a dark wooden branch and looking directly at the camera. Panel 2: A Nubian ibex with long, ribbed, curved horns standing on a rocky outcrop. Panel 3: A golden lion tamarin with vibrant, bright orange fur sitting on a log. Panel 4: An African penguin swimming through clear blue water, viewed from the side. Source: Wildlife Conservation Society
3/ To expand this mission, the Google.org AI for Science fund is funding the sequencing of 150 additional endangered species — like the African penguin and Grevy’s zebra. This data will be openly released for the public and scientists to use in conservation.
A photo of an endangered Grevy’s zebra. Source: Wildlife Conservation Society
A photo of an endangered golden mantella frog. Source: Wildlife Conservation Society
2/ We’re helping @rockefeller.edu and the @ebpgenome.bsky.social preserve the genetic information of every species on Earth. This gives researchers the data they need to prevent further loss of biodiversity.
It once took 13 years and $3 billion to sequence the human genome.
Now, we're using Google’s AI tools to sequence animal genomes in just days to help save endangered species 🧬 (1/4) ↓
goo.gle/4kgx18P
Check out this @wsj.com feature about recent work by the Google Threat Intelligence Group to successfully disrupt IPIDEA, a large residential proxy network that served as the digital camouflage for over 550 threat groups globally. www.wsj.com/tech/google-...
The letter of the day is Y for YouTube: Today, YouTube announced it’s now the largest digital library of "Sesame Street" content. More than 100 full "Sesame Street" episodes from previous seasons are now available on YouTube and the YouTube Kids app for the first time. Learn more → goo.gle/4b2x5qf
4. When it comes to Personal Intelligence in the Gemini app, you are in control.
You can disconnect your apps, adjust your settings, or delete your chat history at any time.
Learn more about how we protect your data → goo.gle/3YBsLH8
3. We’re still learning (and so is Gemini).
We’ve tested this beta extensively, but it might still make mistakes or connect unrelated dots. If Gemini gets it wrong, just tell it.
2. Gemini will try to reference or explain the information it used from your connected sources so you can verify it. If it doesn’t, you can ask it for more information.
1. You decide if and when to connect your apps. This new beta feature is off by default.
You choose which apps to connect, and you can disconnect them anytime.
Today we’re introducing Personal Intelligence in the Gemini app, a new feature that, with your permission, can securely connect info from Google apps like Gmail, Google Photos, Search & YouTube history to make Gemini uniquely helpful.
More on how we built this with user control on the center 🧵↓
Gemini will scan for the imperceptible SynthID watermark across both the audio and visual tracks, specifying which segments contain AI-generated elements.
Learn more ↓ goo.gle/455LSMY
You can now verify Google AI-generated images and videos in the Gemini app ✅
We’re expanding our content transparency tools to help you more easily identify content generated by Google AI. Just upload an image or video file and ask, "Was this generated using Google AI?"
Image
Introducing Gemini 3 Flash, our frontier intelligence model, available at scale for everyone. It excels at coding, tool calling, and is stronger than 2.5 Pro across most metrics!! ⚡️
Available in the API at $0.50 in / 1M tokens and $3.00 out / 1M tokens across. x.com/OfficialLoga...
Gemini 3 Flash is here, bringing a ‘huge’ upgrade to the Gemini app
I’m really excited about our release of Gemini 3 today, the result of hard work by many, many people in the Gemini team and all across Google! 🎊
blog.google/products/gem...
Gemini 3 performs quite well on a wide range of benchmarks.
Ahead of this weekend’s Las Vegas Grand Prix, we’re extending our partnership with the McLaren Formula 1 team, placing Gemini at the heart of the team’s operations. Learn more in Marketing Brew. www.marketingbrew.com/stories/2025...
Gemini 3 has arrived! 🚀
Here are three things to know about our most intelligent model.
Introducing our latest updates to the Gemini app, all built with Gemini 3. It includes new experiments like dynamic view & visual layout that create an interactive interface that’s perfect for your task and Gemini Agent, an experimental tool that can complete multi-step tasks on your behalf.
Learn more about how Gemini 3 can help you learn, build and plan anything → goo.gle/4oUEkVu
We're making Gemini 3 Pro available across the Gemini app and Search, starting with AI Mode. And developers and Google Cloud customers can access Gemini 3 Pro in preview starting today.
Introducing Gemini 3 — our most intelligent model that helps you bring any idea to life.
Gemini 3 has:
🧠 State-of-the-art reasoning
🖼️ Deep multimodal understanding
🤝 Powerful vibe coding and agentic capabilities
