Vivek Natarajan (@vivnat) 's Twitter Profile
Vivek Natarajan

@vivnat

Advancing Biomedical AI @GoogleAI | Previously @facebookai Research | Definite Optimist

ID: 97205267

linkhttps://natviv.me calendar_today16-12-2009 13:11:44

718 Tweet

6,6K Followers

381 Following

DeepLearning.AI (@deeplearningai) 's Twitter Profile Photo

Google presented AI Co-Scientist, a multi-agent system that generates, evaluates, and refines scientific research proposals based on user-defined constraints and ongoing feedback. Using a team of AI agents built on various Gemini 2.0 models, the system has already contributed

Google presented AI Co-Scientist, a multi-agent system that generates, evaluates, and refines scientific research proposals based on user-defined constraints and ongoing feedback. 

Using a team of AI agents built on various Gemini 2.0 models, the system has already contributed
Jeff Dean (@jeffdean) 's Twitter Profile Photo

Our best model*, Gemini 2.5 Pro, is now available for everyone in the Gemini app model drop down menu at gemini.google.com. Give it a try with your most difficult questions! *For now! 😃

Patrick Loeber (@patloeber) 's Twitter Profile Photo

Today marks two months since I joined Google DeepMind, and the pace here is absolutely incredible. Here’s a quick summary of all product updates I got to witness:

Today marks two months since I joined <a href="/GoogleDeepMind/">Google DeepMind</a>, and the pace here is absolutely incredible.

Here’s a quick summary of all product updates I got to witness:
Google DeepMind (@googledeepmind) 's Twitter Profile Photo

When his son Max was diagnosed with a rare disease, our colleague Thomas turned to Gemini to better understand complex scientific papers — and even make new connections across the research community. His story is a powerful reminder of why we work on AI. ↓

Stanford AI+Biomedicine Seminar (@stanford_ai_bio) 's Twitter Profile Photo

We are super excited to have Vivek Natarajan from Google next Tuesday! Vivek is a pioneer in AI + medicine. He is the lead researcher behind the Med-PaLM series and co-led the AI co-scientist and AMIE. He will share with us about his recent work. Join us!

We are super excited to have <a href="/vivnat/">Vivek Natarajan</a> from Google next Tuesday! Vivek is a pioneer in AI + medicine. He is the lead researcher behind the Med-PaLM series and co-led the AI co-scientist and AMIE. He will share with us about his recent work. Join us!
Vivek Natarajan (@vivnat) 's Twitter Profile Photo

Great to see our Google DeepMind Google AI AI co-scientist innovations (multi-agent idea generation tournament) ship fast with Google Cloud Agentspace! Incredible shipping momentum at Google! 🚢 Agentspace -cloud.google.com/products/agent… AI co-scientist - research.google/blog/accelerat…

Alan Karthikesalingam (@alan_karthi) 's Twitter Profile Photo

Happy to share our Google AI Google DeepMind milestone of two nature publications today on Project AMIE - the art of the possible for empathic, conversational, diagnostic medical AI. AMIE exceeded Primary Care Physicians in a randomized comparison of simulated text-based

Anil Palepu (@apalepu13) 's Twitter Profile Photo

Huge thanks to Nathan Labenz for hosting Vivek Natarajan and me on the The Cognitive Revolution Podcast. Really enjoyed discussing Google AI Google DeepMind's latest research on AMIE (A conversational diagnostic AI) & AI co-scientist (AI aiding scientific discovery). youtube.com/watch?v=Jo31LE…

Nathan Labenz (@labenz) 's Twitter Profile Photo

"Dwarkesh Patel should read our papers" Vivek Natarajan & Anil Palepu explain that ... yes, AI systems can generate biomedical hypotheses that human experts find insightful & significant "AI doctors" exist today, and "data centers full of AI geniuses" are pretty clearly coming soon! 🧵

Samuel Hume (@drsamuelbhume) 's Twitter Profile Photo

Top 5 advances in medicine this week (🧵) 1. Doctors vs. AI This is an experiment where patient actors spoke either with a doctor or an AI chatbot (made by Google Research) by typing back and forth - they didn't know which they were speaking to AI won for diagnostic accuracy,

Top 5 advances in medicine this week (🧵)

1. Doctors vs. AI

This is an experiment where patient actors spoke either with a doctor or an AI chatbot (made by Google Research) by typing back and forth - they didn't know which they were speaking to 

AI won for diagnostic accuracy,
👩‍💻 Paige Bailey (@dynamicwebpaige) 's Twitter Profile Photo

Favorite image from the Google DeepMind Co-Scientist paper, showing how generative AI can accelerate hypothesis development and experimentation. Work that previously would have taken years or a decade, can now be accomplished in just a matter of days with Gemini 2.5 and beyond:

Favorite image from the <a href="/GoogleDeepMind/">Google DeepMind</a> Co-Scientist paper, showing how generative AI can accelerate hypothesis development and experimentation.

Work that previously would have taken years or a decade, can now be accomplished in just a matter of days with Gemini 2.5 and beyond:
Jessica Sacher, PhD (@jessicasacher) 's Twitter Profile Photo

Does every scientist need an AI co-scientist? We sat down with José R Penadés & Tiago Costa, two Imperial College London professors putting @google's new 'AI co-scientist' tool to the test. How long did it take AI to solve the mystery they'd spent a decade trying to figure out? 🧵

Does every scientist need an AI co-scientist?

We sat down with <a href="/jrpenades/">José R Penadés</a> &amp; <a href="/CostaT_Lab/">Tiago Costa</a>, two <a href="/imperialcollege/">Imperial College London</a> professors putting @google's new 'AI co-scientist' tool to the test.

How long did it take AI to solve the mystery they'd spent a decade trying to figure out? 🧵
Vivek Natarajan (@vivnat) 's Twitter Profile Photo

Out-the-box gemini 2.0 flash in an agentic framework surpasses doctors in multimodal medical conservations. This is a pretty big deal for medicine. Huge congrats to @KhaledSaab11 Ryu Tanno | 丹野 龍太郎 Jan Freyberg + the whole AMIE team Google DeepMind Google AI on this accomplishment!

Ryu Tanno | 丹野 龍太郎 (@ryutarotanno) 's Twitter Profile Photo

Excited to share a big update on AMIE, our research AI doctor from Google DeepMind and Google AI Now, AMIE can “see” and interpret visual medical data within a diagnostic conversation. Yes, AMIE exceeded human doctors (PCPs) in many key metrics like diagnostic accuracy and

Excited to share a big update on AMIE, our research AI doctor from <a href="/GoogleDeepMind/">Google DeepMind</a> and <a href="/GoogleAI/">Google AI</a> 

Now, AMIE can “see” and interpret visual medical data within a diagnostic conversation.

Yes, AMIE exceeded human doctors (PCPs) in many key metrics like diagnostic accuracy and
Ethan Mollick (@emollick) 's Twitter Profile Photo

Study in Nature: “Across 30 out of 32 evaluation axes from the specialist physician perspective & 25 out of 26 evaluation axes from the patient-actor perspective, AMIE [Google Medical LLM] was rated superior to PCPs [primary care physicians] while being non-inferior on the rest.”

Study in Nature: “Across 30 out of 32 evaluation axes from the specialist physician perspective &amp; 25 out of 26 evaluation axes from the patient-actor perspective, AMIE [Google Medical LLM] was rated superior to PCPs [primary care physicians] while being non-inferior on the rest.”
Vivek Natarajan (@vivnat) 's Twitter Profile Photo

Excited to share new validation results of our Google DeepMind Google AI AI co-scientist! Addressing the challenge of liver fibrosis, a condition affecting millions globally with few treatments options, we've been working with Dr. Gary Peltz Stanford Medicine, exploring how

Vivek Natarajan (@vivnat) 's Twitter Profile Photo

Congrats to Google DeepMind teammates on the stunning AlphaEvolve work! With AI co-scientist and AlphaEvolve, we are now seeing the emergence of language-guided search and optimization with reasoning LLMs as a new paradigm, distinct and perhaps complementary to RL /