AlphaGo
- AlphaGo. Google Deepmind
- PUBLICATION Super-Exponential Regret for UCT, AlphaGo and Variants
- PROJECT AlphaGo
- RESEARCH AlphaGo Zero: Starting from scratch
- RESEARCH AlphaGo’s next move
- RESEARCH Innovations of AlphaGo
- RESEARCH Exploring the mysteries of Go with AlphaGo and China’s top players
- RESPONSIBILITY & SAFETY Taking a responsible path to AGI
- RESEARCH Genie 2: A large-scale foundation world model
- PROJECT AlphaZero and MuZero
- RESEARCH How AlphaChip transformed computer chip design
- THE PODCAST AI and neuroscience: The virtuous circle
- THE PODCAST Go to Zero
- IMPACT AlphaDev discovers faster sorting algorithms
- IMPACT MuZero’s first step from research into the real world
- RESEARCH AlphaFold: Using AI for scientific discovery
- RESEARCH MuZero: Mastering Go, chess, shogi and Atari without rules
- RESEARCH Specification gaming: the flip side of AI ingenuity
- RESEARCH AlphaFold: Using AI for scientific discovery
- COMPANY From unlikely start-up to major scientific organisation: Entering our tenth year at DeepMind
- RESEARCH AlphaStar: Grandmaster level in StarCraft II using multi-agent reinforcement learning
- COMPANY The Podcast: Episode 2: Go to Zero
- COMPANY Welcome to the DeepMind podcast
- RESEARCH TF-Replicator: Distributed Machine Learning for Researchers
- RESEARCH AlphaZero: Shedding new light on chess, shogi, and Go
- RESEARCH Game-theory insights into asymmetric multi-agent games
- COMPANY 2017: DeepMind’s year in review
- RESEARCH Agents that imagine and plan
- COMPANY DeepMind’s work in 2016: a round-up
- RESEARCH Decoupled Neural Interfaces Using Synthetic Gradients
- RESEARCH Deep Reinforcement Learning
AlphaZero
- PROJECT AlphaZero
- PROJECT AlphaZero and MuZero
- PUBLICATION Finding Increasingly Large Extremal Graphs with AlphaZero and Tabu Search
- IMPACT MuZero, AlphaZero, and AlphaDev: Optimizing computer systems
- RESEARCH AlphaZero: Shedding new light on chess, shogi, and Go
- PUBLICATION Super-Exponential Regret for UCT, AlphaGo and Variants
- PUBLICATION Self-Predictive Universal AI
- PROJECT AlphaGo
- RESEARCH How AlphaChip transformed computer chip design
- RESEARCH AI achieves silver-medal standard solving International Mathematical Olympiad problems
- THE PODCAST Go to Zero
- IMPACT AlphaDev discovers faster sorting algorithms
- RESEARCH Mastering Stratego, the classic game of imperfect information
- RESEARCH Discovering novel algorithms with AlphaTensor
- RESEARCH DeepMind’s latest research at ICLR 2022
- IMPACT MuZero’s first step from research into the real world
- RESEARCH AlphaFold: Using AI for scientific discovery
- RESEARCH Generally capable agents emerge from open-ended play
- RESEARCH MuZero: Mastering Go, chess, shogi and Atari without rules
- RESEARCH AlphaFold: Using AI for scientific discovery
- RESEARCH AlphaStar: Grandmaster level in StarCraft II using multi-agent reinforcement learning
- COMPANY The Podcast: Episode 2: Go to Zero
- RESEARCH AlphaStar: Mastering the real-time strategy game StarCraft II
AlphaFold
- TECHNOLOGY AlphaFold Database
- TECHNOLOGY AlphaFold Server
- RESEARCH AlphaFold 3 predicts the structure and interactions of all of life’s molecules
- RESEARCH A glimpse of the next generation of AlphaFold
- TECHNOLOGY AlphaFold
- COMPANY How our principles helped define AlphaFold’s release
- IMPACT AlphaFold transforms biology for millions around the world
- IMPACT AlphaFold unlocks one of the greatest puzzles in biology
- RESEARCH AlphaFold reveals the structure of the protein universe
- RESEARCH Putting the power of AlphaFold into the world’s hands
- RESEARCH AlphaFold: Using AI for scientific discovery
- RESEARCH AlphaFold: a solution to a 50-year-old grand challenge in biology
- RESEARCH AlphaFold: Using AI for scientific discovery
- PUBLICATION Accurate proteome-wide missense variant effect prediction with AlphaMissense
- RESEARCH NeurIPS 2024
- COMPANY Demis Hassabis & John Jumper awarded Nobel Prize in Chemistry
- TECHNOLOGY Impact stories
- IMPACT STORIES Creating a protein structure database
- IMPACT STORIES Breaking down plastic pollution
- IMPACT STORIES Accelerating the fight against malaria
- IMPACT STORIES Using AI to spot osteoporosis earlier
- IMPACT STORIES Understanding faulty proteins
- IMPACT STORIES Paving the way for potential Parkinson’s treatments
- IMPACT STORIES Solving one of biology’s greatest puzzles
- IMPACT STORIES Designing more effective drugs
- IMPACT STORIES Increasing honeybees’ chances of survival
- IMPACT STORIES Searching for life-changing medicines
- IMPACT STORIES Tracing proteins back to origins of life
- IMPACT STORIES Racing against drug-resistant bacteria
- IMPACT STORIES Solving an extinction mystery
- RESEARCH AlphaProteo generates novel proteins for biology and health research
- RESPONSIBILITY & SAFETY Looking ahead to the AI Seoul Summit
- RESEARCH NeurIPS 2022
- RESEARCH ICLR 2023
- RESEARCH ICML 2023
- COMPANY 2023: A Year of Groundbreaking Advances in AI and Computing
- ABOUT Responsibility & Safety
- DISCOVER Visualising AI
- THE PODCAST Out of the lab
- THE PODCAST AI and neuroscience: The virtuous circle
- THE PODCAST A breakthrough unfolds
- RESEARCH A catalogue of genetic mutations to help pinpoint the cause of diseases
- RESEARCH Google DeepMind’s latest research at ICML 2023
- IMPACT Google Cloud: Driving digital transformation
- RESEARCH Benchmarking the next generation of never-ending learners
- IMPACT Stopping malaria in its tracks
- RESEARCH Measuring perception in AI models
- IMPACT Fighting osteoporosis before it starts
- IMPACT Understanding the faulty proteins linked to cancer and autism
- IMPACT Solving the mystery of how an ancient bird went extinct
- IMPACT Targeting early-onset Parkinson’s with AI
- COMPANY Maximising the impact of our breakthroughs
- IMPACT The race to cure a billion people from a deadly parasitic disease
- IMPACT Tracing the evolution of proteins back to the origin of life
- IMPACT How the honeybee could help protect species around the world
- IMPACT Advancing discovery of better drugs and medicine
- IMPACT Creating plastic-eating enzymes that could save us from pollution
- IMPACT Accelerating the race against antibiotic resistance
- RESEARCH Enabling high-accuracy protein structure prediction at the proteome scale
- RESEARCH Computational predictions of protein structures associated with COVID-19
- COMPANY The Podcast: Episode 5: Out of the lab
- RESEARCH TF-Replicator: Distributed Machine Learning for Researchers
- RESEARCH AlphaStar: Mastering the real-time strategy game StarCraft II
- RESEARCH AlphaZero: Shedding new light on chess, shogi, and Go
AlphaTensor
- PROJECT AlphaTensor
- PUBLICATION AlphaTensor for Optimizing Quantum Computations
- RESEARCH Discovering novel algorithms with AlphaTensor
- RESEARCH AlphaEvolve: A Gemini-powered coding agent for designing advanced algorithms
- RESEARCH NeurIPS 2022
- RESEARCH ICLR 2023
AlphaDev
- PROJECT AlphaDev
- IMPACT MuZero, AlphaZero, and AlphaDev: Optimizing computer systems
- IMPACT AlphaDev discovers faster sorting algorithms
- PROJECT AlphaGo
- COMPANY 2023: A Year of Groundbreaking Advances in AI and Computing
AlphaGeometry
AlphaEvolve
- AlphaEvolve: A coding agent for scientific and algorithmic discovery. Google DeepMind
- RESEARCH AlphaEvolve: A Gemini-powered coding agent for designing advanced algorithms
New AI agent evolves algorithms for math and practical applications in computing by combining the creativity of large language models with automated evaluators
Large language models (LLMs) are remarkably versatile. They can summarize documents, generate code or even brainstorm new ideas. And now we’ve expanded these capabilities to target fundamental and highly complex problems in mathematics and modern computing.
Today, we’re announcing AlphaEvolve, an evolutionary coding agent powered by large language models for general-purpose algorithm discovery and optimization. AlphaEvolve pairs the creative problem-solving capabilities of our Gemini modelswith automated evaluators that verify answers, and uses an evolutionary framework to improve upon the most promising ideas.
AlphaEvolve enhanced the efficiency of Google’s data centers, chip design and AI training processes — including training the large language models underlying AlphaEvolve itself. It has also helped design faster matrix multiplication algorithms and find new solutions to open mathematical problems, showing incredible promise for application across many areas.
- Designing better algorithms with large language models
- Optimizing our computing ecosystem
- Improving data center scheduling
- Assisting in hardware design
- Enhancing AI training and inference
- Advancing the frontiers in mathematics and algorithm discover