From conducting fundamental research to influencing product development, our research teams have the opportunity to impact technology used by billions of people every day. We aspire to make discoveries that impact everyone, and sharing our research and tools to fuel progress in the field is fundamental to our approach.
External link for Google Research
Google Research reposted this
Scientists are transforming how we understand diseases, from forecasting flu surges to detecting cancer earlier. Through collaborative research and investment, Google is supporting these teams in moving critical discoveries from the lab to the clinic. Learn more about how we're helping put better health within reach for everyone: https://lnkd.in/eM6AxcKs.
Can two researchers and 200,000 iterations of code out-predict traditional epidemiological models? 📊🩺 In this clip from The Google Research Podcast, host Sinéad Bovell sits down with Google Research’s Michael Brenner and Lizzie Dorfman to discuss how their AI model achieved the CDC’s most accurate forecasts for flu and RSV hospitalizations in 2025–26. Key takeaways from the discussion: 1️⃣ Scale Through Iteration Achieving high-precision forecasting required running over 200,000 code variations — demonstrating how rapid AI experimentation can refine complex biological parameters faster than traditional methods. 2️⃣ Actionable Public Health Intelligence Predictive modeling isn't just about data points; it’s about giving hospitals and public health officials the advance lead time needed to allocate critical resources before surge capacity is reached. 3️⃣ Small Teams, High Impact A streamlined team leveraging advanced machine learning tools can tackle massive epidemiological challenges, unlocking new paradigms for how public health research is conducted. 👉 Watch the full episode here: https://lnkd.in/gJK3yJy8 Explore more stories from Google Research: https://goo.gle/3NvR4kt
Translating conversational diagnostic AI from simulation into real clinical workflows requires rigorous research. Today in The Lancet Group, Google, and Beth Israel Deaconess Medical Center share results from a prospective feasibility study evaluating AMIE ahead of urgent primary care visits: - Real-time Oversight: Seven board-certified internal medicine physicians monitored 100 live patient interactions, resulting in zero protocol-defined safety interruptions. - Diagnostic Reasoning: AMIE's differential matched the adjudicated reference diagnosis in 75% of cases within the top 3 candidates and 90% within the top 7. - Workflow Integration: Masked clinical evaluators rated AMIE's communication and history-gathering quality favorably across 17 clinical criteria, with strong reception from clinicians preparing for consultations. This milestone provides a foundation for future multi-arm clinical studies assessing our research medical system in real-world clinical settings and how it could improve patient-doctor relationships. Explore the complete research and methodology: https://lnkd.in/g2uQrNAf
Interested in privacy-preserving synthetic data? Catch Alex Bie at the @COLM_conf Google booth (#107) today, at 2:00 PM PT for an encore presentation of ContinuousBench, evaluating knowledge transfer and data contamination in DP synthesis.@GoogleDeepMind Join the conversation at #COLM2026!
Join Alex Bie at the Conference on Language Modeling (COLM 2026) Google booth #107 today at 5:00 PM for a walkthrough of ContinuousBench, a standardized benchmark designed to measure knowledge transfer in differentially private (DP) synthesis. Don't miss the chance to explore if DP synthetic data truly preserve information, or just style? @GoogleDeepMind Read the paper: https://lnkd.in/gRyMirj9
Today at 12:00pm, don't miss our live demonstration on EmbeddingGemma 2 at the @COLM_conf Google booth (#107). Connect with Sahil Dua and explore an open multimodal model that unifies text, images, audio, and video representations. #COLM2026 @GoogleDeepMind
Missed yesterday's demo on adaptive agent environments? Stop by the #COLM2026 Google booth #107 today at 11:00 AM PT to catch Zifeng Wang presenting EnvHarness — a flexible framework enabling co-evolution between LLM agents and their training environments.
LLM agents learn by interacting with environments, but static setups limit their growth. Today at 2:00 PM, join Zifeng Wang at the #COLM2026 Google booth (#107) to learn about EnvHarness, a plug-in architecture that dynamically reshapes environment behaviors to improve reinforcement learning and agent adaptability.
Join Alex Bie at the Conference on Language Modeling (COLM 2026) Google booth #107 today at 5:00 PM for a walkthrough of ContinuousBench, a standardized benchmark designed to measure knowledge transfer in differentially private (DP) synthesis. Don't miss the chance to explore if DP synthetic data truly preserve information, or just style? @GoogleDeepMind Read the paper: https://lnkd.in/gRyMirj9
Today we announce new findings from Visiting Fellow David Autor on how AI impacts how professionals build expertise. In a three-month randomized controlled trial with practicing patent attorneys, we test both short-term productivity & longer-term skill building that occur as a result of AI usage. More: goo.gle/4ASXC3R
LLM agents learn by interacting with environments, but static setups limit their growth. Today at 2:00 PM, join Zifeng Wang at the #COLM2026 Google booth (#107) to learn about EnvHarness, a plug-in architecture that dynamically reshapes environment behaviors to improve reinforcement learning and agent adaptability.
Can autonomous AI advance the frontiers of scientific discovery? Join Rui Meng at the @COLM_conf Google booth (#107) today at 2:00 PM PT for a live demo of ScientistTwo, an autonomous multi-agent framework that analyzes research papers, identifies limitations, and produces verified codebases. Read the paper: https://lnkd.in/gFMgN7nT