正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Excited to announce AlphaEvolve A powerful AI coding agent developed by our team in Google DeepMind that is able to discover impactful new algorithms for important problems in Maths and Computing by combining the creativity of large language models with automated evaluators.

Pushmeet Kohli

25,586 subscribers

203,414 次观看 • 1 年前 •via X (Twitter)

科学技术新闻政治教育

Anya Rossi• Live Now

Private livecam show

11 条评论

Pushmeet Kohli 的头像

Pushmeet Kohli1 年前

We’ve deployed algorithms that AlphaEvolve discovered in key areas across Google: 1️⃣To make data centres more efficient - saving Google’s worldwide compute resources 2️⃣To improve TPU chip design 3️⃣To enhance AI training of Gemini - the model underpinning AlphaEvolve itself

Pushmeet Kohli 的头像

Pushmeet Kohli1 年前

AlphaEvolve discovered new efficient algorithm for matrix multiplication, a fundamental problem in Computer Science, improving Strassen's method that has stood for 50 years. Applied to 50 open Math problems, it matched the best known answers 75% of the time and improved on 20%.

Pushmeet Kohli 的头像

Pushmeet Kohli1 年前

It’s been incredible to see the range of complex problems AlphaEvolve can tackle and excited to see what we can take on next. Read more:

NICE 的头像

NICE1 年前

Stay competitive by balancing cutting-edge AI with automation tools. Forrester shows how.

Oriol Vinyals 的头像

Oriol Vinyals1 年前

@GoogleDeepMind It's overwhelming (in a good way) to see all the many way in which @Google is embracing and developing cool stuff on top of Gemini. Congrats!

Aharon Azulay 的头像

Aharon Azulay1 年前

@GoogleDeepMind Recursive self improvement alert

Adhil Parammel 的头像

Adhil Parammel1 年前

@GoogleDeepMind Kudos to you and team. This is history in making.

Ketan Singh 的头像

Ketan Singh1 年前

@GoogleDeepMind Hey @grok @AskPerplexity, please share a comprehensive list of what this is capable of, which wasn't easily possible before it.

Singularity 🪙 的头像

Singularity 🪙1 年前

@GoogleDeepMind AGI/ASI COMİNG !!!!!!🚀🚀🚀🚀

rohit 的头像

rohit1 年前

@GoogleDeepMind Once again, may I say, please make some of these truly incredible tech breakthroughs make its way into product!

RobKai 🇩🇪 的头像

RobKai 🇩🇪1 年前

@GoogleDeepMind This is an important step to AGI 🔥

相关视频

New No Priors with Pushmeet Kohli Matej Balog from DeepMind AlphaEvolve on: *ai co-scientists *improving matrix multiplication *reclaiming ~1% of google compute *can AI discover new knowledge? *evolutionary algorithms

New No Priors with Pushmeet Kohli Matej Balog from DeepMind AlphaEvolve on: ai co-scientists improving matrix multiplication reclaiming ~1% of google compute can AI discover new knowledge? *evolutionary algorithms

sarah guo

17,590 次观看 • 11 个月前

Google DeepMind, David Silver reveals: we built a system that used RL to discover its own RL algorithms. this AI-designed system outperformed all human-created RL algorithms developed over the years.

Google DeepMind, David Silver reveals: we built a system that used RL to discover its own RL algorithms. this AI-designed system outperformed all human-created RL algorithms developed over the years.

Haider.

397,103 次观看 • 1 年前

🚀 Excited to announce the first release of a novel open source programming language and platform for language model interaction! Combining prompts, constraints & scripting, LMQL elevates the capabilities of large language models. 🧵1/6 A quick tour.

🚀 Excited to announce the first release of a novel open source programming language and platform for language model interaction! Combining prompts, constraints & scripting, LMQL elevates the capabilities of large language models. 🧵1/6 A quick tour.

LMQL (Language Model Query Language)

198,966 次观看 • 3 年前

What if LLMs could find fundamentally new solutions to hard problems? AlphaEvolve is an evolutionary coding agent built on top of Gemini for scientific discoveries. 🧵👇 🔵 Matrix Multiplication Re-invented (0:00) 🔵 AlphaEvolve's Architecture (3:08) 🔵 Math to Google's Core (3:37) 🔵 Spectrum of Discovery (4:21) 🔵 Creativity (5:14, 7:11) 🔵 Representational Power (6:26) 🔵 Self-Improving Systems (8:18) 🔵 Value of Programmatic Discovery (8:46) 🔵 Synergy: AlphaEvolve & LLMs (11:58) 🔵 Managing Expensive Validations (14:53) 🔵 Future: "Library of Algorithms" (17:17) Clips from our interview with Google DeepMind researchers Matej Balog and Alexander Novikov on AlphaEvolve.

What if LLMs could find fundamentally new solutions to hard problems? AlphaEvolve is an evolutionary coding agent built on top of Gemini for scientific discoveries. 🧵👇 🔵 Matrix Multiplication Re-invented (0:00) 🔵 AlphaEvolve's Architecture (3:08) 🔵 Math to Google's Core (3:37) 🔵 Spectrum of Discovery (4:21) 🔵 Creativity (5:14, 7:11) 🔵 Representational Power (6:26) 🔵 Self-Improving Systems (8:18) 🔵 Value of Programmatic Discovery (8:46) 🔵 Synergy: AlphaEvolve & LLMs (11:58) 🔵 Managing Expensive Validations (14:53) 🔵 Future: "Library of Algorithms" (17:17) Clips from our interview with Google DeepMind researchers Matej Balog and Alexander Novikov on AlphaEvolve.

Machine Learning Street Talk

11,104 次观看 • 1 年前

Today Google DeepMind released AlphaEvolve: a Gemini coding agent for algorithm discovery. It beat the famous Strassen algorithm for matrix multiplication set 56 years ago. Google has been killing it recently. We had early access to the paper and interviewed the researchers.

Today Google DeepMind released AlphaEvolve: a Gemini coding agent for algorithm discovery. It beat the famous Strassen algorithm for matrix multiplication set 56 years ago. Google has been killing it recently. We had early access to the paper and interviewed the researchers.

Machine Learning Street Talk

58,495 次观看 • 1 年前

Google DeepMind CEO Demis Hassabis says AI agents are coming that will be able to search the Tree of Knowledge to discover new and original things about the world

Google DeepMind CEO Demis Hassabis says AI agents are coming that will be able to search the Tree of Knowledge to discover new and original things about the world

Tsarathustra

42,910 次观看 • 1 年前

1M users joined our community! Thanks for your love! Now we are excited to announce that AI Terminal is ready to launch. Here is a 5 minutes live demo without any edit. AI Terminal is the early version of Agent browser (you can also call it Agent OS). It's backed by our agent network routing and orchestration technologies. Chainopera AI is built by crypto, for crypto! We are looking forwards to see great agents and AI models contributed by the community. We will officially release it to our early community remembers starting from ETH Denver. See you in our booth!

1M users joined our community! Thanks for your love! Now we are excited to announce that AI Terminal is ready to launch. Here is a 5 minutes live demo without any edit. AI Terminal is the early version of Agent browser (you can also call it Agent OS). It's backed by our agent network routing and orchestration technologies. Chainopera AI is built by crypto, for crypto! We are looking forwards to see great agents and AI models contributed by the community. We will officially release it to our early community remembers starting from ETH Denver. See you in our booth!

ChainOpera AI

550,737 次观看 • 1 年前

Today Google Maps is getting its biggest upgrade in over a decade. By combining our Gemini models with a deep understanding of the world, Maps now unlocks entirely new possibilities for how you navigate and explore. Here’s what you need to know 🧵

Today Google Maps is getting its biggest upgrade in over a decade. By combining our Gemini models with a deep understanding of the world, Maps now unlocks entirely new possibilities for how you navigate and explore. Here’s what you need to know 🧵

Google

28,834,283 次观看 • 3 个月前

"We are entering a new era of coding." Nobel Prize Winner and Google DeepMind CEO Demis Hassabis on coding with natural language.

"We are entering a new era of coding." Nobel Prize Winner and Google DeepMind CEO Demis Hassabis on coding with natural language.

Reid Hoffman

175,956 次观看 • 1 年前

We created a concept album about AI with award-winning musicians and composers for #GoogleCloudNext and used the Music AI Sandbox, developed by Google DeepMind, and tools from Google Cloud to bring it to life. Press ▶️ to learn more ↓

We created a concept album about AI with award-winning musicians and composers for #GoogleCloudNext and used the Music AI Sandbox, developed by Google DeepMind, and tools from Google Cloud to bring it to life. Press ▶️ to learn more ↓

Google Cloud

30,319 次观看 • 1 年前

We are incredibly excited to announce River AI. Our mission is to create personal AI that is owned and shaped by you. Today’s best AIs are controlled by a few large corporations. We are building the alternative: a new, personal stack for AI that works entirely for you, shares your values, and operates on your terms.

We are incredibly excited to announce River AI. Our mission is to create personal AI that is owned and shaped by you. Today’s best AIs are controlled by a few large corporations. We are building the alternative: a new, personal stack for AI that works entirely for you, shares your values, and operates on your terms.

River AI

738,733 次观看 • 7 天前

New AI Package with DeepSeek AI Model TesseractAI continues to expand and improve our Custom AI Agent. You will now be able to add AI packages to your agent in addition to custom information. 🌟 This new feature is being developed using #DeepSeek AI model. In particular, DeepSeek R1 is designed for greater efficiency and more goal-oriented tasks, requiring fewer computational resources while still delivering powerful results. Additionally, we’re experimenting with DeepSeek’s technology for our existing Agent LLM to further enhance performance. Our goal is to make the Custom AI Agent even more efficient and effective for all your needs.

New AI Package with DeepSeek AI Model TesseractAI continues to expand and improve our Custom AI Agent. You will now be able to add AI packages to your agent in addition to custom information. 🌟 This new feature is being developed using #DeepSeek AI model. In particular, DeepSeek R1 is designed for greater efficiency and more goal-oriented tasks, requiring fewer computational resources while still delivering powerful results. Additionally, we’re experimenting with DeepSeek’s technology for our existing Agent LLM to further enhance performance. Our goal is to make the Custom AI Agent even more efficient and effective for all your needs.

Tesseract AI

13,236 次观看 • 1 年前

Demis Hassabis clarifies that while DeepMind is exploring powerful AI systems, the goal is not to trigger an uncontrolled intelligence explosion. Instead, his team is experimenting with combining evolutionary programming and foundation models to explore how self-improving systems might evolve—carefully and responsibly.

Demis Hassabis clarifies that while DeepMind is exploring powerful AI systems, the goal is not to trigger an uncontrolled intelligence explosion. Instead, his team is experimenting with combining evolutionary programming and foundation models to explore how self-improving systems might evolve—carefully and responsibly.

Wes Roth

19,194 次观看 • 1 年前

Introducing ALE-Bench, ALE-Agent! Towards Automating Long-Horizon Algorithm Engineering for Hard Optimization Problems Blog: Paper: ALE-Bench is a coding benchmark primarily focused on hard optimization (NP-hard) problems. We developed this benchmark with AtCoder Inc., a leading coding contest platform company. What makes ALE-Bench unique is its focus on hard optimization problems that demand long-horizon and creative reasoning. It’s open-ended, in the sense that true optima are out of reach (NP-hard) and scores can continuously improve. We believe this benchmark has the potential to become one of the key benchmarks for reasoning and coding in the next generation. ALE-Agent is our end-to-end agent that we specifically designed for this challenging domain. In fact, our ALE-Agent has already built an impressive track record in the wild! In May 2025, our agent participated in a live AtCoder Heuristic Competition (AHC), alongside 1,000 other participants in real-time. AHC is considered to be one of the most challenging coding competitions in this domain. Our ALE-Agent achieved an impressive ranking of 21st out of 1,000 human participants in the competition (top 2%), marking a turning point for AI discovery of solutions to hard optimization problems with a wide spectrum of important real world applications such as logistics, routing, packing, factory production planning, power-grid balancing. We look forward to applying this technology to real industrial optimization opportunities. Building on the insights from this study, Sakana AI will continue to tackle the challenge of developing AI with even greater algorithm engineering capabilities. ALE-Bench Dataset: ALE-Bench Code: This research was conducted in collaboration with AtCoder Inc. (AtCoder). We are deeply grateful for their outstanding expertise and contributions in optimization and algorithms, which were invaluable in providing data, analyzing results, and enabling our AI agent’s participation in their contests.

Introducing ALE-Bench, ALE-Agent! Towards Automating Long-Horizon Algorithm Engineering for Hard Optimization Problems Blog: Paper: ALE-Bench is a coding benchmark primarily focused on hard optimization (NP-hard) problems. We developed this benchmark with AtCoder Inc., a leading coding contest platform company. What makes ALE-Bench unique is its focus on hard optimization problems that demand long-horizon and creative reasoning. It’s open-ended, in the sense that true optima are out of reach (NP-hard) and scores can continuously improve. We believe this benchmark has the potential to become one of the key benchmarks for reasoning and coding in the next generation. ALE-Agent is our end-to-end agent that we specifically designed for this challenging domain. In fact, our ALE-Agent has already built an impressive track record in the wild! In May 2025, our agent participated in a live AtCoder Heuristic Competition (AHC), alongside 1,000 other participants in real-time. AHC is considered to be one of the most challenging coding competitions in this domain. Our ALE-Agent achieved an impressive ranking of 21st out of 1,000 human participants in the competition (top 2%), marking a turning point for AI discovery of solutions to hard optimization problems with a wide spectrum of important real world applications such as logistics, routing, packing, factory production planning, power-grid balancing. We look forward to applying this technology to real industrial optimization opportunities. Building on the insights from this study, Sakana AI will continue to tackle the challenge of developing AI with even greater algorithm engineering capabilities. ALE-Bench Dataset: ALE-Bench Code: This research was conducted in collaboration with AtCoder Inc. (AtCoder). We are deeply grateful for their outstanding expertise and contributions in optimization and algorithms, which were invaluable in providing data, analyzing results, and enabling our AI agent’s participation in their contests.

Sakana AI

237,195 次观看 • 1 年前

All remote cognitive jobs, done using only a laptop setup, is most exposed to be automated or heavily reduced by AI. ~ Shane Legg, Chief AGI Scientist and co-founder of Google DeepMind ---- Full video from 'Google DeepMind and Hannah Fry' YT channel (link in comment)

All remote cognitive jobs, done using only a laptop setup, is most exposed to be automated or heavily reduced by AI. ~ Shane Legg, Chief AGI Scientist and co-founder of Google DeepMind ---- Full video from 'Google DeepMind and Hannah Fry' YT channel (link in comment)

Rohan Paul

81,268 次观看 • 5 个月前

Demis Hassabis says hallucination isn't always a bug. It can be a tool for creativity. To solve hard problems with AlphaEvolve, DeepMind forced the models to hallucinate. Most ideas were nonsense but some turned out to be useful. "You can substitute the word hallucination for imagination at that point. They're obviously two sides of the same coin."

Demis Hassabis says hallucination isn't always a bug. It can be a tool for creativity. To solve hard problems with AlphaEvolve, DeepMind forced the models to hallucinate. Most ideas were nonsense but some turned out to be useful. "You can substitute the word hallucination for imagination at that point. They're obviously two sides of the same coin."

vitrupo

309,802 次观看 • 1 年前

OpenAI Co-founder Andrej Karpathy explains the new computing paradigm: "We're entering a new computing paradigm with large language models acting like CPUs, using tokens instead of bytes, and having a context window instead of RAM. This is the Large Language Model OS (LMOS)"

OpenAI Co-founder Andrej Karpathy explains the new computing paradigm: "We're entering a new computing paradigm with large language models acting like CPUs, using tokens instead of bytes, and having a context window instead of RAM. This is the Large Language Model OS (LMOS)"

Haider.

552,309 次观看 • 1 年前

We're excited to unveil a new leadership team for Ogilvy One and a fresh approach to designing customer relationships – one led by creativity, powered by data, and supercharged by AI. Learn more:

We're excited to unveil a new leadership team for Ogilvy One and a fresh approach to designing customer relationships – one led by creativity, powered by data, and supercharged by AI. Learn more:

Ogilvy

10,815 次观看 • 2 年前

With AI, "we did 1,000,000,000 years of PHD time in one year." Nobel Prize Winner and Google DeepMind CEO Demis Hassabis on AlphaFold, the groundbreaking AI system developed to solve one of biology’s most complex problems: protein folding.

With AI, "we did 1,000,000,000 years of PHD time in one year." Nobel Prize Winner and Google DeepMind CEO Demis Hassabis on AlphaFold, the groundbreaking AI system developed to solve one of biology’s most complex problems: protein folding.

Reid Hoffman

113,442 次观看 • 1 年前

Here we go: Gemini 3.0 is live! Thankfully, I was able to be an early tester. It's the best model I've ever tested - nothing comes even close to it. OpenAI will have a very hard time to keep up with Google, because Gemini 3.0 will now be the benchmark for all future models. I have been invited by Google DeepMind / Google - Thanks to Google DeepMind for the #EarlyAccess Here's a summary of Gemini 3.0 with all the important benchmark evals:

Here we go: Gemini 3.0 is live! Thankfully, I was able to be an early tester. It's the best model I've ever tested - nothing comes even close to it. OpenAI will have a very hard time to keep up with Google, because Gemini 3.0 will now be the benchmark for all future models. I have been invited by Google DeepMind / Google - Thanks to Google DeepMind for the #EarlyAccess Here's a summary of Gemini 3.0 with all the important benchmark evals:

Chubby♨️

108,611 次观看 • 7 个月前