Google DeepMind (@GoogleDeepMind) / X

Google DeepMind

4,438 posts

Google DeepMind

@GoogleDeepMind

The engine room of @Google. Building AI safely and responsibly to solve the world’s most complex problems. Join us: deepmind.google/careers/

London, UK

Joined January 2016

Pinned
Google DeepMind
@GoogleDeepMind
22h
DiffusionGemma is our new experimental open model with up to 4x faster output on dedicated GPUs. Instead of predicting word-by-word, it generates entire blocks of text simultaneously. This lets the model self-correct and format complex markdown in real time.
145K
Google DeepMind
@GoogleDeepMind
3h
When millions of AI agents interact with each other, new collective behaviors can emerge. 🌐 Together with @schmidtsciences, @coop_ai, @ARIA_research and supported by @Googleorg, we’re launching a $10M research fund to help understand how AI systems behave as a group. →
GIF
24K
Google DeepMind
@GoogleDeepMind
19h
Replying to @GoogleDeepMind
We evaluated AI’s impact by looking beyond test scores to behavioral shifts. Over eight weeks, results suggest students were using AI to understand concepts, not just find answers – with Gemini queries about how to tackle problems rising from 68% to 90%. Find out more →
Gemini’s guided learning: results from a randomized controlled trial in Sierra Leone
From deepmind.google
16K
Google DeepMind
@GoogleDeepMind
19h
In Sierra Leone, a surging student population is outpacing available teachers. Our latest research explores how AI can act as a partner to support educators in these environments – amplifying their reach without replacing their essential expertise and skills. 🧵
GIF
23K
Google DeepMind
@GoogleDeepMind
22h
Replying to @GoogleDeepMind
Find out more → goo.gle/4vG0xcI
19K
Google DeepMind
@GoogleDeepMind
Jun 9
Replying to @GoogleDeepMind
Try it in the @Google Translate app or build in preview via the API in @GoogleAIStudio. Find out more ↓
Fluid, natural voice translation with Gemini 3.5 Live Translate
From blog.google
19K
Google DeepMind
@GoogleDeepMind
Jun 9
Replying to @GoogleDeepMind
3.5 Live Translate can convert speech into over 70 languages and processes it as it’s streamed - while keeping tone, pace and pitch intact - allowing for more natural conversations.
00:00
27K
Google DeepMind
@GoogleDeepMind
Jun 9
Say hello, hola, 你好 to Gemini 3.5 Live Translate: our latest audio model built for fast, cross-language communication. 🌐
GIF
121K
Google DeepMind reposted
Google Gemma
@googlegemma
Jun 3
Meet Gemma 4 12B! A unified, encoder-free multimodal model designed to bring high-performance intelligence directly to your laptop, and released under an Apache 2.0 license. Bridging the gap between edge efficiency and advanced reasoning. Here is what’s new with Gemma 4 12B: 👇
3.2M
Google DeepMind
@GoogleDeepMind
Jun 2
Replying to @GoogleDeepMind
We’re excited to be making Co-Scientist available to individual researchers through Hypothesis Generation as part of Gemini for Science: a collection of experiments and tools exploring the future of AI-powered scientific discovery. Find out more →
Co-Scientist: A multi-agent AI partner to accelerate research
From deepmind.google
19K
Google DeepMind
@GoogleDeepMind
Jun 2
Replying to @GoogleDeepMind
Over the past year, we’ve collaborated with global scientific experts to evaluate the system on complex problems. It assisted by: identifying new targets for liver fibrosis, uncovering fresh approaches to tackling Amyotrophic lateral sclerosis (ALS), and digested decades of
21K
Google DeepMind
@GoogleDeepMind
Jun 2
Replying to @GoogleDeepMind
Co-Scientist can produce thousands of hypotheses. To find the most useful, it relies on a “tournament of ideas” and holds scientific debates to refine and rank these. It can also verify claims against scientific literature and data, and use web search and specialized models to
GIF
13K
Google DeepMind
@GoogleDeepMind
Jun 2
Replying to @GoogleDeepMind
Scientific discovery is a cycle of ideation, critique, and refinement. Co-Scientist mirrors this using a coalition of specialized Gemini-based agents leveraging its reasoning, multimodal, long-context and tool-use capabilities.
15K