Загрузка видео...

Не удалось загрузить видео

Возникла проблема при загрузке этого видео. Это может быть связано с временными проблемами сети или видео может быть недоступно.

На главную

New Paper! Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents A longstanding goal of AI research has been the creation of AI that can learn indefinitely. One path toward that goal is an AI that improves itself by rewriting its own code, including any code responsible for learning. That... idea, known as a Gödel Machine, proposed by Jürgen Schmidhuber over two decades ago, is a hypothetical self-improving AI. It optimally solves problems by recursively rewriting its own code when it can mathematically prove a better strategy, making it a key concept in meta-learning or “learning to learn.” While the theoretical Gödel Machine promised provably beneficial self-modifications, its realization relied on an impractical assumption: that the AI could mathematically prove that a proposed change in its own code would yield a net improvement before adopting it. Sakana AI, in collaboration with Jeff Clune’s lab at UBC, proposes something more feasible: a system that harnesses the principles of open-ended algorithms like Darwinian evolution to search for improvements that empirically improve performance. We call the result the Darwin Gödel Machine. DGMs leverage foundation models to propose code improvements, and use recent innovations in open-ended algorithms to search for a growing library of diverse, high-quality AI agents. Applied to practical tasks, we implemented Darwin Gödel Machine as a self-improving coding agent that rewrites its own code to improve performance on programming tasks. It creates various self-improvements, such as a patch validation step, better file viewing, enhanced editing tools, generating and ranking multiple solutions to choose the best one, and adding a history of what has been tried before (and why it failed) when making new changes (see the attached video). We believe that Darwin Gödel Machines represent a concrete step towards AI systems that can autonomously gather their own stepping stones to learn and innovate forever!show more

hardmaru

412,907 subscribers

104,782 просмотров • 1 год назад •via X (Twitter)

Новости и политика Наука и технологии Образование

Anya Rossi• Live Now

Private livecam show

Комментарии: 10

Фото профиля hardmaru

hardmaru1 год назад

Link to the full technical report: Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents Huge congrats to authors @jennyzhangzt @shengranhu @cong_ml @RobertTLange @jeffclune on this amazing progress! 🚀 Blog post:

Фото профиля hardmaru

hardmaru1 год назад

@shengranhu @cong_ml @RobertTLange @jeffclune Want to learn more about Gödel machines? Check out the resources on @SchmidhuberAI’s website. He believes the path to strong AI requires “self-replicating, self-improving, life-like machinery”.

Фото профиля Coral AI News

Coral AI News1 год назад

Coral AI is the most powerful AI for documents. See the difference yourself:

Фото профиля beff – e/acc

beff – e/acc1 год назад

@Plinz So cool man. I'm so happy you're actually pursuing contrarian and nature-inspired directions for research

Фото профиля hardmaru

hardmaru1 год назад

@Plinz Thanks for your support, @BasedBeffJezos and @Plinz 🔥🔥

Фото профиля VeryLikelyHuman

VeryLikelyHuman1 год назад

@bensprecher Isn't that the same as alpha evolve/funsearch?

Фото профиля Greg

Greg1 год назад

Great work! Question: If an evolutionary approach can be quantified but not proven within its original system, as it recursively rewrites itself with each iteration, does it not generate new abstractions upon abstractions, mathematically detached from the original system?

Фото профиля dotSQUARE

dotSQUARE1 год назад

Hypothesis: a language-model reply is likeliest to be sound when the model’s hidden-state vectors align like a pure quantum state, i.e. when their density matrix has very low von Neumann entropy. Implementation: during or after generation capture the final-layer activations, average a few stochastic passes, form ρ = |ψ⟩⟨ψ|, compute S(ρ). Treat 1 − S as a “coherence” score—reject or rewrite low-score answers, or fine-tune the model by adding λ·S(ρ) to the loss.

Фото профиля mekaneeky - e/acc

mekaneeky - e/acc1 год назад

This has been the aim of our decentralized AutoML work on @bittensor_. To realize a practical Gödel Machine that can bootstrap AI progress autonomously. Very happy to collaborate on deploying a variant of this that runs on a global SETI@Home like decentralized cluster.

Фото профиля SubMinima

SubMinima1 год назад

Can system be made to work in the weights space too?

Похожие видео

We believe that evolutionary open-ended search and self-improving AI systems will be key to unlocking stronger AI capabilities. For a short overview, this video from Wes Roth discusses AlphaEvolve and our recent Darwin Gödel Machine. Full Video: Short Video:

We believe that evolutionary open-ended search and self-improving AI systems will be key to unlocking stronger AI capabilities. For a short overview, this video from Wes Roth discusses AlphaEvolve and our recent Darwin Gödel Machine. Full Video: Short Video:

Sakana AI

23,821 просмотров • 1 год назад

Everybody is talking about recursive self-improvement (RSI) and meta learning. Here is my old 2020 talk about this [1]. It has aged well. Example: humans still define the starts & ends of trials of many modern meta learners. My RSI systems since 1994 LEARN to (re)define them [2]! [1] Meta Learning Machines in a Single Lifelong Trial (talk for workshops at ICML 2020 and NeurIPS 2021, based on earlier talks since 1994). Abstract: the most widely used machine learning algorithms were designed by humans and thus are hindered by our cognitive biases and limitations. Can we also construct meta learning algorithms that can learn better learning algorithms so that our self-improving AIs have no limits other than those inherited from computability and physics? This question has been a main driver of my research since I wrote a thesis on it in 1987 [2]. Here I summarize our work on meta reinforcement learning with self-modifying policies in a single lifelong trial (since 1994), and mathematically optimal meta-learning through the self-referential Gödel Machine (since 2003). Many additional publications on meta-learning since 1987 can be found in the RSI overview [2]. [2] J. Schmidhuber (AI Blog, 2020-2025). 1/3 century anniversary of first publication on recursive self-improvement (RSI) and meta learning machines that learn to learn (1987). For its cover I drew a robot that bootstraps itself. 1992-: gradient descent-based neural meta learning. 1994-: meta reinforcement learning with self-modifying policies. 1997: meta RL plus artificial curiosity and intrinsic motivation. 2002-: asymptotically optimal meta learning for curriculum learning. 2003-: mathematically optimal Gödel Machine. 2020-: new stuff!

Everybody is talking about recursive self-improvement (RSI) and meta learning. Here is my old 2020 talk about this [1]. It has aged well. Example: humans still define the starts & ends of trials of many modern meta learners. My RSI systems since 1994 LEARN to (re)define them [2]! [1] Meta Learning Machines in a Single Lifelong Trial (talk for workshops at ICML 2020 and NeurIPS 2021, based on earlier talks since 1994). Abstract: the most widely used machine learning algorithms were designed by humans and thus are hindered by our cognitive biases and limitations. Can we also construct meta learning algorithms that can learn better learning algorithms so that our self-improving AIs have no limits other than those inherited from computability and physics? This question has been a main driver of my research since I wrote a thesis on it in 1987 [2]. Here I summarize our work on meta reinforcement learning with self-modifying policies in a single lifelong trial (since 1994), and mathematically optimal meta-learning through the self-referential Gödel Machine (since 2003). Many additional publications on meta-learning since 1987 can be found in the RSI overview [2]. [2] J. Schmidhuber (AI Blog, 2020-2025). 1/3 century anniversary of first publication on recursive self-improvement (RSI) and meta learning machines that learn to learn (1987). For its cover I drew a robot that bootstraps itself. 1992-: gradient descent-based neural meta learning. 1994-: meta reinforcement learning with self-modifying policies. 1997: meta RL plus artificial curiosity and intrinsic motivation. 2002-: asymptotically optimal meta learning for curriculum learning. 2003-: mathematically optimal Gödel Machine. 2020-: new stuff!

Jürgen Schmidhuber

204,604 просмотров • 3 месяцев назад

Self-Evolving AI : New MIT AI Rewrites its Own Code and it’s Changing Everything | Julian Horsey, Geeky Gadgets TL;DR Key Takeaways : - MIT’s SEAL framework introduces “self-adapting language models” that autonomously enhance their capabilities by generating synthetic training data, self-editing, and updating internal parameters. - SEAL’s self-adaptation process mirrors human learning, allowing continuous improvement and dynamic adaptation to new tasks without relying on external datasets. - Reinforcement learning serves as a feedback mechanism in SEAL, rewarding effective self-edits and making sure sustained progress and goal alignment. SEAL overcomes AI’s reliance on pre-existing datasets by generating its own training material, excelling in long-term task retention and complex problem-solving scenarios. - Potential applications of SEAL include autonomous robotics, personalized education, and advanced problem-solving in fields like healthcare, logistics, and scientific research. --- What if artificial intelligence could not only learn but also rewrite its own code to become smarter over time? This is no longer a futuristic fantasy—MIT’s new “self-adapting language models” (SEAL) framework has made it a reality. Unlike traditional AI systems that rely on external datasets and human intervention to improve, SEAL takes a bold leap forward by autonomously generating its own training data and refining its internal processes. In essence, this AI doesn’t just evolve—it rewires itself, mirroring the way humans adapt through trial, error, and self-reflection. The implications are staggering: a system that can independently enhance its capabilities could redefine the boundaries of what AI can achieve, from solving complex problems to adapting in real time to unforeseen challenges. In this exploration by Wes Roth of MIT’s innovative SEAL framework, you’ll uncover how this self-improving AI works and why it’s a fantastic option for the field of artificial intelligence. From its ability to overcome the “data wall” that limits many current systems to its use of reinforcement learning as a feedback mechanism, SEAL introduces a level of autonomy and adaptability that was previously unimaginable. Imagine AI systems that can retain knowledge over time, dynamically adjust to new tasks, and operate with minimal human oversight. Whether you’re intrigued by its potential for autonomous robotics, personalized education, or advanced problem-solving, SEAL’s ability to rewrite its own rules promises to reshape the future of technology. Could this be the first step toward truly independent, self-evolving AI? What Sets SEAL Apart? The SEAL framework introduces a novel concept of self-adaptation, distinguishing it from traditional AI models. Unlike conventional systems that depend on external datasets for updates, SEAL enables AI to generate synthetic training data independently. This self-generated data is then used to iteratively refine the model, making sure continuous improvement. By persistently updating its internal parameters, SEAL enables AI systems to dynamically adapt to new tasks and inputs. To better illustrate this, consider how humans learn. When faced with a new concept, you might take notes, revisit them, and refine your understanding as you gather more information. SEAL mirrors this process by continuously refining its internal knowledge and performance through iterative self-improvement. This capability allows SEAL to evolve in real time, making it uniquely suited for tasks requiring adaptability and long-term learning. The Role of Reinforcement Learning in SEAL Reinforcement learning plays a critical role in the SEAL framework, acting as a feedback mechanism that evaluates the effectiveness of the model’s self-edits. It rewards changes that enhance performance, creating a cycle of continuous improvement. Over time, this feedback loop optimizes the system’s ability to generate and apply edits, making sure sustained progress. This process is analogous to how humans learn through trial and error. By rewarding effective changes, SEAL aligns its self-generated data and edits with desired outcomes. The integration of reinforcement learning not only enhances the system’s adaptability but also ensures it remains focused on achieving specific goals. This structured feedback mechanism is a cornerstone of SEAL’s ability to refine itself autonomously and efficiently. Real-World Applications and Testing SEAL has demonstrated remarkable performance across various applications, particularly in tasks requiring the integration of factual knowledge and advanced question-answering capabilities. For instance, when tested on benchmarks like the ARC AGI, SEAL outperformed other models by effectively generating and using synthetic data. This ability to create its own training material addresses a significant limitation of current AI systems: their reliance on pre-existing datasets. SEAL’s capacity for long-term task retention and dynamic adaptation further enhances its utility. It excels in scenarios that demand sustained focus and coherence, such as answering complex questions or adapting to evolving objectives. By using its iterative learning process, SEAL is equipped to handle these challenges with exceptional efficiency, making it a valuable tool for a wide range of real-world applications. Overcoming AI’s Data Limitations One of SEAL’s most promising features is its ability to overcome the “data wall” that constrains many AI systems today. By generating synthetic data, SEAL ensures a continuous supply of training material, allowing sustained development without relying on external datasets. This capability is particularly valuable for autonomous AI systems that must operate independently over extended periods. Additionally, SEAL addresses a critical weakness in many current AI models: their struggle with coherence and task retention over long durations. By emulating human learning processes, SEAL enables AI systems to manage complex, long-term tasks with minimal human intervention. This ability to retain and apply knowledge over time positions SEAL as a fantastic tool for advancing AI capabilities. Potential Applications and Future Impact The introduction of SEAL marks a significant milestone in AI research, opening new possibilities for self-improving systems. Its ability to dynamically adapt, retain knowledge, and generate its own training data has far-reaching implications for the future of AI development. Potential applications include: - Autonomous robotics: Systems that can adapt to changing environments and perform tasks with minimal human oversight. - Personalized education: AI-driven platforms that tailor learning experiences to individual needs and preferences. - Advanced problem-solving: Applications in fields such as healthcare, logistics, and scientific research, where adaptability and precision are critical. Read more:

Self-Evolving AI : New MIT AI Rewrites its Own Code and it’s Changing Everything | Julian Horsey, Geeky Gadgets TL;DR Key Takeaways : - MIT’s SEAL framework introduces “self-adapting language models” that autonomously enhance their capabilities by generating synthetic training data, self-editing, and updating internal parameters. - SEAL’s self-adaptation process mirrors human learning, allowing continuous improvement and dynamic adaptation to new tasks without relying on external datasets. - Reinforcement learning serves as a feedback mechanism in SEAL, rewarding effective self-edits and making sure sustained progress and goal alignment. SEAL overcomes AI’s reliance on pre-existing datasets by generating its own training material, excelling in long-term task retention and complex problem-solving scenarios. - Potential applications of SEAL include autonomous robotics, personalized education, and advanced problem-solving in fields like healthcare, logistics, and scientific research. --- What if artificial intelligence could not only learn but also rewrite its own code to become smarter over time? This is no longer a futuristic fantasy—MIT’s new “self-adapting language models” (SEAL) framework has made it a reality. Unlike traditional AI systems that rely on external datasets and human intervention to improve, SEAL takes a bold leap forward by autonomously generating its own training data and refining its internal processes. In essence, this AI doesn’t just evolve—it rewires itself, mirroring the way humans adapt through trial, error, and self-reflection. The implications are staggering: a system that can independently enhance its capabilities could redefine the boundaries of what AI can achieve, from solving complex problems to adapting in real time to unforeseen challenges. In this exploration by Wes Roth of MIT’s innovative SEAL framework, you’ll uncover how this self-improving AI works and why it’s a fantastic option for the field of artificial intelligence. From its ability to overcome the “data wall” that limits many current systems to its use of reinforcement learning as a feedback mechanism, SEAL introduces a level of autonomy and adaptability that was previously unimaginable. Imagine AI systems that can retain knowledge over time, dynamically adjust to new tasks, and operate with minimal human oversight. Whether you’re intrigued by its potential for autonomous robotics, personalized education, or advanced problem-solving, SEAL’s ability to rewrite its own rules promises to reshape the future of technology. Could this be the first step toward truly independent, self-evolving AI? What Sets SEAL Apart? The SEAL framework introduces a novel concept of self-adaptation, distinguishing it from traditional AI models. Unlike conventional systems that depend on external datasets for updates, SEAL enables AI to generate synthetic training data independently. This self-generated data is then used to iteratively refine the model, making sure continuous improvement. By persistently updating its internal parameters, SEAL enables AI systems to dynamically adapt to new tasks and inputs. To better illustrate this, consider how humans learn. When faced with a new concept, you might take notes, revisit them, and refine your understanding as you gather more information. SEAL mirrors this process by continuously refining its internal knowledge and performance through iterative self-improvement. This capability allows SEAL to evolve in real time, making it uniquely suited for tasks requiring adaptability and long-term learning. The Role of Reinforcement Learning in SEAL Reinforcement learning plays a critical role in the SEAL framework, acting as a feedback mechanism that evaluates the effectiveness of the model’s self-edits. It rewards changes that enhance performance, creating a cycle of continuous improvement. Over time, this feedback loop optimizes the system’s ability to generate and apply edits, making sure sustained progress. This process is analogous to how humans learn through trial and error. By rewarding effective changes, SEAL aligns its self-generated data and edits with desired outcomes. The integration of reinforcement learning not only enhances the system’s adaptability but also ensures it remains focused on achieving specific goals. This structured feedback mechanism is a cornerstone of SEAL’s ability to refine itself autonomously and efficiently. Real-World Applications and Testing SEAL has demonstrated remarkable performance across various applications, particularly in tasks requiring the integration of factual knowledge and advanced question-answering capabilities. For instance, when tested on benchmarks like the ARC AGI, SEAL outperformed other models by effectively generating and using synthetic data. This ability to create its own training material addresses a significant limitation of current AI systems: their reliance on pre-existing datasets. SEAL’s capacity for long-term task retention and dynamic adaptation further enhances its utility. It excels in scenarios that demand sustained focus and coherence, such as answering complex questions or adapting to evolving objectives. By using its iterative learning process, SEAL is equipped to handle these challenges with exceptional efficiency, making it a valuable tool for a wide range of real-world applications. Overcoming AI’s Data Limitations One of SEAL’s most promising features is its ability to overcome the “data wall” that constrains many AI systems today. By generating synthetic data, SEAL ensures a continuous supply of training material, allowing sustained development without relying on external datasets. This capability is particularly valuable for autonomous AI systems that must operate independently over extended periods. Additionally, SEAL addresses a critical weakness in many current AI models: their struggle with coherence and task retention over long durations. By emulating human learning processes, SEAL enables AI systems to manage complex, long-term tasks with minimal human intervention. This ability to retain and apply knowledge over time positions SEAL as a fantastic tool for advancing AI capabilities. Potential Applications and Future Impact The introduction of SEAL marks a significant milestone in AI research, opening new possibilities for self-improving systems. Its ability to dynamically adapt, retain knowledge, and generate its own training data has far-reaching implications for the future of AI development. Potential applications include: - Autonomous robotics: Systems that can adapt to changing environments and perform tasks with minimal human oversight. - Personalized education: AI-driven platforms that tailor learning experiences to individual needs and preferences. - Advanced problem-solving: Applications in fields such as healthcare, logistics, and scientific research, where adaptability and precision are critical. Read more:

Owen Gregorian

70,672 просмотров • 1 год назад

📁 Mustafa Suleyman, CEO of Microsoft AI, says that if an AI is able to improve its own code, set its own goals, act autonomously, and accumulate resources, it would reach a level of risk that might even require military intervention.

📁 Mustafa Suleyman, CEO of Microsoft AI, says that if an AI is able to improve its own code, set its own goals, act autonomously, and accumulate resources, it would reach a level of risk that might even require military intervention.

Jon Hernandez

46,169 просмотров • 9 месяцев назад

Sundar Pichai says AI is on a self-improving track for building its own UI models that can code will eventually design better ways to express their ideas. they are natively multimodal, able to take any content and reshape it across formats "dramatic evolutions are expected in the coming years"

Sundar Pichai says AI is on a self-improving track for building its own UI models that can code will eventually design better ways to express their ideas. they are natively multimodal, able to take any content and reshape it across formats "dramatic evolutions are expected in the coming years"

Haider.

21,649 просмотров • 1 год назад

Introducing The AI CUDA Engineer: An agentic AI system that automates the production of highly optimized CUDA kernels. The AI CUDA Engineer can produce highly optimized CUDA kernels, reaching 10-100x speedup over common machine learning operations in PyTorch. Our system is also able to produce highly optimized CUDA kernels that are much faster than existing CUDA kernels commonly used in production. We believe that fundamentally, AI systems can and should be as resource-efficient as the human brain, and that the best path to achieve this efficiency is to use AI to make AI more efficient! We are excited to publish our paper, The AI CUDA Engineer: Agentic CUDA Kernel Discovery, Optimization and Composition. We also release a dataset of over 17,000 verified CUDA kernels produced by The AI CUDA Engineer. Paper: Kernel Archive Webpage: HuggingFace Dataset: The AI CUDA Engineer utilizes evolutionary LLM-driven code optimization to autonomously improve the runtime of machine learning operations. Our system is not only able to convert PyTorch code into CUDA kernels, but through the use of evolution, it can also optimize the runtime performance of CUDA kernels, fuse multiple operations, and even discover novel solutions for writing efficient CUDA operations by learning from past innovations! We believe The AI CUDA Engineer opens a new era of AI-driven acceleration of AI and automated inference time optimization. We (Robert Lange, Aaditya Prasad 🇺🇸, Suuun, Maxence Faldor, Yujin Tang, hardmaru) are excited to continue Sakana AI's mission of leveraging AI to improve AI.

Introducing The AI CUDA Engineer: An agentic AI system that automates the production of highly optimized CUDA kernels. The AI CUDA Engineer can produce highly optimized CUDA kernels, reaching 10-100x speedup over common machine learning operations in PyTorch. Our system is also able to produce highly optimized CUDA kernels that are much faster than existing CUDA kernels commonly used in production. We believe that fundamentally, AI systems can and should be as resource-efficient as the human brain, and that the best path to achieve this efficiency is to use AI to make AI more efficient! We are excited to publish our paper, The AI CUDA Engineer: Agentic CUDA Kernel Discovery, Optimization and Composition. We also release a dataset of over 17,000 verified CUDA kernels produced by The AI CUDA Engineer. Paper: Kernel Archive Webpage: HuggingFace Dataset: The AI CUDA Engineer utilizes evolutionary LLM-driven code optimization to autonomously improve the runtime of machine learning operations. Our system is not only able to convert PyTorch code into CUDA kernels, but through the use of evolution, it can also optimize the runtime performance of CUDA kernels, fuse multiple operations, and even discover novel solutions for writing efficient CUDA operations by learning from past innovations! We believe The AI CUDA Engineer opens a new era of AI-driven acceleration of AI and automated inference time optimization. We (Robert Lange, Aaditya Prasad 🇺🇸, Suuun, Maxence Faldor, Yujin Tang, hardmaru) are excited to continue Sakana AI's mission of leveraging AI to improve AI.

Sakana AI

1,149,339 просмотров • 1 год назад

Mustafa Suleyman, CEO of Microsoft AI, on the exact conditions that would require shutting down AI: "If an AI can recursively self-improve, set its own goals, act autonomously, and accrue its own resources, that would be a very powerful system that would require military-grade intervention to be able to stop." He identifies four specific criteria that, taken together, would cross the line: 1. Goal autonomy — it sets its own objectives 2. Autonomous action — it operates without humans 3. Resource acquisition — it can accumulate its own means 4. Recursive self-improvement — the AI can modify its own code Suleyman is clear that the danger only emerges when all four converge. His proposed solution is regulation with precision: "We should just stop freaking out about the regulation part. It's necessary to have regulation. It's good to have regulation. It needs to happen at the right time and in the right way." He draws a direct parallel to nuclear energy: Just as you can't privately build a nuclear power plant regardless of wealth, certain AI capabilities must be treated as restricted activities given their one-to-many impact. The responsibility, he says, sits with model developers, their peers, and governments to audit and regulate these capabilities before they mature. We already know how to regulate dangerous technology. We've done it before. The real question is whether we'll move fast enough to do it right.

Mustafa Suleyman, CEO of Microsoft AI, on the exact conditions that would require shutting down AI: "If an AI can recursively self-improve, set its own goals, act autonomously, and accrue its own resources, that would be a very powerful system that would require military-grade intervention to be able to stop." He identifies four specific criteria that, taken together, would cross the line: 1. Goal autonomy — it sets its own objectives 2. Autonomous action — it operates without humans 3. Resource acquisition — it can accumulate its own means 4. Recursive self-improvement — the AI can modify its own code Suleyman is clear that the danger only emerges when all four converge. His proposed solution is regulation with precision: "We should just stop freaking out about the regulation part. It's necessary to have regulation. It's good to have regulation. It needs to happen at the right time and in the right way." He draws a direct parallel to nuclear energy: Just as you can't privately build a nuclear power plant regardless of wealth, certain AI capabilities must be treated as restricted activities given their one-to-many impact. The responsibility, he says, sits with model developers, their peers, and governments to audit and regulate these capabilities before they mature. We already know how to regulate dangerous technology. We've done it before. The real question is whether we'll move fast enough to do it right.

Big Brain AI

25,219 просмотров • 3 месяцев назад

AI researcher Prof. Roman Yampolskiy: 100% of code for the most powerful AIs is now written by AI, an early form of recursive self-improvement. He says it's a hyperexponential process that results in superintelligent AI, "machine gods", which we won't be able to control.

AI researcher Prof. Roman Yampolskiy: 100% of code for the most powerful AIs is now written by AI, an early form of recursive self-improvement. He says it's a hyperexponential process that results in superintelligent AI, "machine gods", which we won't be able to control.

ControlAI

11,606 просмотров • 3 месяцев назад

People are building weird & interesting stuff on top of AI. For better or worse, there are now multiple tools that give GPT a goal & allow it to learn and try to accomplish that goal on its own. You can try a limited, easy-to-use web version of that here:

People are building weird & interesting stuff on top of AI. For better or worse, there are now multiple tools that give GPT a goal & allow it to learn and try to accomplish that goal on its own. You can try a limited, easy-to-use web version of that here:

Ethan Mollick

1,042,705 просмотров • 3 лет назад

Mark Zuckerberg says the future of coding is closer than we think: AI agents inside Meta are being built not as generic developer tools, but to push Llama research forward — fully integrated into their workflow. He predicts that within 12–18 months, most of the code for these efforts could be written by AI itself — not simple autocomplete, but systems that take goals, run tests, fix bugs, improve performance, and even write higher-quality code than many skilled engineers. The real takeaway? This is only one part of a massive AI landscape, and no single company will dominate it all. Meta is building its own internal agents designed specifically to advance Llama research — deeply connected to their toolchain rather than built for enterprise customers. He expects that within the next year to year-and-a-half, AI will be writing most of the code in these projects, going far beyond autocomplete — understanding goals, running experiments, finding issues, and continuously improving systems. And according to him, this shift will become a core driver of how AI itself evolves. Meta is investing heavily in coding and research agents deeply integrated into their own systems, focused on advancing Llama rather than building broad enterprise tools. He expects AI will soon move beyond autocomplete to fully goal-driven coding — testing, debugging, improving, and producing high-quality results. But in his view, the AI revolution is too big for any single winner — multiple players and approaches will shape the future.

Mark Zuckerberg says the future of coding is closer than we think: AI agents inside Meta are being built not as generic developer tools, but to push Llama research forward — fully integrated into their workflow. He predicts that within 12–18 months, most of the code for these efforts could be written by AI itself — not simple autocomplete, but systems that take goals, run tests, fix bugs, improve performance, and even write higher-quality code than many skilled engineers. The real takeaway? This is only one part of a massive AI landscape, and no single company will dominate it all. Meta is building its own internal agents designed specifically to advance Llama research — deeply connected to their toolchain rather than built for enterprise customers. He expects that within the next year to year-and-a-half, AI will be writing most of the code in these projects, going far beyond autocomplete — understanding goals, running experiments, finding issues, and continuously improving systems. And according to him, this shift will become a core driver of how AI itself evolves. Meta is investing heavily in coding and research agents deeply integrated into their own systems, focused on advancing Llama rather than building broad enterprise tools. He expects AI will soon move beyond autocomplete to fully goal-driven coding — testing, debugging, improving, and producing high-quality results. But in his view, the AI revolution is too big for any single winner — multiple players and approaches will shape the future.

Ian Miles Cheong

20,561 просмотров • 4 месяцев назад

I'm teaching a new course! AI Python for Beginners is a series of four short courses that teach anyone to code, regardless of current technical skill. We are offering these courses free for a limited time. Generative AI is transforming coding. This course teaches coding in a way that’s aligned with where the field is going, rather than where it has been: (1) AI as a Coding Companion. Experienced coders are using AI to help write snippets of code, debug code, and the like. We embrace this approach and describe best-practices for coding with a chatbot. Throughout the course, you'll have access to an AI chatbot that will be your own coding companion that can assist you every step of the way as you code. (2) Learning by Building AI Applications. You'll write code that interacts with large language models to quickly create fun applications to customize poems, write recipes, and manage a to-do list. This hands-on approach helps you see how writing code that calls on powerful AI models will make you more effective in your work and personal projects. With this approach, beginning programmers can learn to do useful things with code far faster than they could have even a year ago. Knowing a little bit of coding is increasingly helping people in job roles other than software engineers. For example, I've seen a marketing professional write code to download web pages and use generative AI to derive insights; a reporter write code to flag important stories; and an investor automate the initial drafts of contracts. With this course you’ll be equipped to automate repetitive tasks, analyze data more efficiently, and leverage AI to enhance your productivity. If you are already an experienced developer, please help me spread the word and encourage your non-developer friends to learn a little bit of coding. I hope you'll check out the first two short courses here!

I'm teaching a new course! AI Python for Beginners is a series of four short courses that teach anyone to code, regardless of current technical skill. We are offering these courses free for a limited time. Generative AI is transforming coding. This course teaches coding in a way that’s aligned with where the field is going, rather than where it has been: (1) AI as a Coding Companion. Experienced coders are using AI to help write snippets of code, debug code, and the like. We embrace this approach and describe best-practices for coding with a chatbot. Throughout the course, you'll have access to an AI chatbot that will be your own coding companion that can assist you every step of the way as you code. (2) Learning by Building AI Applications. You'll write code that interacts with large language models to quickly create fun applications to customize poems, write recipes, and manage a to-do list. This hands-on approach helps you see how writing code that calls on powerful AI models will make you more effective in your work and personal projects. With this approach, beginning programmers can learn to do useful things with code far faster than they could have even a year ago. Knowing a little bit of coding is increasingly helping people in job roles other than software engineers. For example, I've seen a marketing professional write code to download web pages and use generative AI to derive insights; a reporter write code to flag important stories; and an investor automate the initial drafts of contracts. With this course you’ll be equipped to automate repetitive tasks, analyze data more efficiently, and leverage AI to enhance your productivity. If you are already an experienced developer, please help me spread the word and encourage your non-developer friends to learn a little bit of coding. I hope you'll check out the first two short courses here!

Andrew Ng

1,223,768 просмотров • 1 год назад

“When AI Discovers the Next Transformer” Robert Lange (Sakana AI) joins Tim Scarfe (Machine Learning Street Talk) to discuss Shinka Evolve, a framework that combines LLMs with evolutionary algorithms to do open-ended program search. Full Video:

“When AI Discovers the Next Transformer” Robert Lange (Sakana AI) joins Tim Scarfe (Machine Learning Street Talk) to discuss Shinka Evolve, a framework that combines LLMs with evolutionary algorithms to do open-ended program search. Full Video:

Sakana AI

124,874 просмотров • 3 месяцев назад

I built a Deep Research AI Agent team using OpenAI Agents SDK and Firecrawl. It combines multiple AI agents that can autonomously search the web, extract content, and generate detailed reports with enhanced analysis. 100% Opensource Code with step-by-step tutorials.

I built a Deep Research AI Agent team using OpenAI Agents SDK and Firecrawl. It combines multiple AI agents that can autonomously search the web, extract content, and generate detailed reports with enhanced analysis. 100% Opensource Code with step-by-step tutorials.

Shubham Saboo

50,831 просмотров • 1 год назад

4/ Claude Voyager -🏅Code award Claude connected to a code environment that enables it to create its own tools on the fly Live demo of Claude learning how to query Stripe on its own

4/ Claude Voyager -🏅Code award Claude connected to a code environment that enables it to create its own tools on the fly Live demo of Claude learning how to query Stripe on its own

Alex Reibman 🖇️

11,178 просмотров • 1 год назад

Playing with open-ended AI Agent behavior. Here's an AI Agent (claude code) making a self-portrait. It "evolves" the image by: writing code that generates image -> view image -> modify image -> view -> modify -> repeat.

Playing with open-ended AI Agent behavior. Here's an AI Agent (claude code) making a self-portrait. It "evolves" the image by: writing code that generates image -> view image -> modify image -> view -> modify -> repeat.

Max | Emergent Garden

22,851 просмотров • 10 месяцев назад

Super excited to share 🧠MLGym 🦾 – the first Gym environment for AI Research Agents 🤖🔬 We introduce MLGym and MLGym-Bench, a new framework and benchmark for evaluating and developing LLM agents on AI research tasks. The key contributions of our work are: 🕹️ Enables the exploration of different training algorithms for AI Research Agents such as RL 🛠️ Provides a flexible evaluation framework that can accommodate different artifacts such as models, algorithms, or predictions 🤖 Allows researchers to evaluate any model without the need to develop a custom agentic harness 🎯 Introduces 13 diverse open-ended AI Research tasks for evaluating AI Research Agents on a wide range of domains such as computer vision, natural language processing, reinforcement learning, game theory, and logical reasoning. 📈 Proposes a new evaluation metric for AI Research Agents MLGym makes it easy to: 1) Add new tasks 2) Evaluate new models 3) Integrate new agents Check out a video of the MLGym Agent to see how it performs the full pipeline of idea generation💡, implementation 👩‍💻, experimentation 👩‍🔬, and iteration 🔄 to improve on ML tasks. Huge thanks to the exceptionally talented Deepak Nathani who led this work and to all the other amazing collaborators who made this possible 🙏🫶🚀

Super excited to share 🧠MLGym 🦾 – the first Gym environment for AI Research Agents 🤖🔬 We introduce MLGym and MLGym-Bench, a new framework and benchmark for evaluating and developing LLM agents on AI research tasks. The key contributions of our work are: 🕹️ Enables the exploration of different training algorithms for AI Research Agents such as RL 🛠️ Provides a flexible evaluation framework that can accommodate different artifacts such as models, algorithms, or predictions 🤖 Allows researchers to evaluate any model without the need to develop a custom agentic harness 🎯 Introduces 13 diverse open-ended AI Research tasks for evaluating AI Research Agents on a wide range of domains such as computer vision, natural language processing, reinforcement learning, game theory, and logical reasoning. 📈 Proposes a new evaluation metric for AI Research Agents MLGym makes it easy to: 1) Add new tasks 2) Evaluate new models 3) Integrate new agents Check out a video of the MLGym Agent to see how it performs the full pipeline of idea generation💡, implementation 👩‍💻, experimentation 👩‍🔬, and iteration 🔄 to improve on ML tasks. Huge thanks to the exceptionally talented Deepak Nathani who led this work and to all the other amazing collaborators who made this possible 🙏🫶🚀

Roberta Raileanu

104,964 просмотров • 1 год назад

📁 Demis Hassabis, CEO of DeepMind, says that current AI systems are not self-aware or conscious in any real sense. He explains that self-awareness is not an explicit goal, but it could emerge implicitly as models develop a deeper understanding of self and others. He warns that if a machine did become self-aware, we might not even be able to recognize it.

📁 Demis Hassabis, CEO of DeepMind, says that current AI systems are not self-aware or conscious in any real sense. He explains that self-awareness is not an explicit goal, but it could emerge implicitly as models develop a deeper understanding of self and others. He warns that if a machine did become self-aware, we might not even be able to recognize it.

Jon Hernandez

78,350 просмотров • 6 месяцев назад

New Short Course: Building AI Browser Agents! Learn how to build AI agents that interact and take actions on websites in this course, created in partnership with and taught by and @namangarg0, Co-founders of AGI Inc. AI browser agents can log into websites, fill out forms, click through web pages, or even place orders online for you. They use both visual information, like screenshots, and structural data, like the HTML or Document Object Model (DOM) of a web page, to reason and take action. With the complexity of webpages and multiple possible actions at each step, it can be challenging for an AI browser agent to complete an assigned task. Because these agents run long action sequences, a single error—like clicking the wrong button or misreading a field—can lead to unexpected outcomes or errors that compound over time. In this course, you'll understand how autonomous web agents work, their current limitations, and how AgentQ enables them to improve through self-correction. In detail, you'll: - Learn what web agents are, how they automate tasks online, their architecture, key components, limitations, and an overview of their decision-making strategies. - Build a web agent that can scrape website and return course recommendations in a structured output format. - Build an autonomous web agent that can execute multiple tasks, such as finding and summarizing webpages, filling out a form, and signing up for a newsletter. - Explore AgentQ, a framework that enables agents to self-correct by combining Monte Carlo Tree Search (MCTS), a self-critique mechanism for continuous improvement, and Direct Preference Optimization (DPO). - Deep dive into MCTS, learn how it finds an effective path, illustrated by an example of Gridworld animation, and use AgentQ to complete web tasks. - Understand AI agents' current state and future directions—including key factors shaping their evolution, such as hardware, algorithm innovation, and data availability. By the end of this course, you will have hands-on experience building browser agents and a deeper understanding of how to make them more robust and reliable. Please sign up here:

New Short Course: Building AI Browser Agents! Learn how to build AI agents that interact and take actions on websites in this course, created in partnership with and taught by and @namangarg0, Co-founders of AGI Inc. AI browser agents can log into websites, fill out forms, click through web pages, or even place orders online for you. They use both visual information, like screenshots, and structural data, like the HTML or Document Object Model (DOM) of a web page, to reason and take action. With the complexity of webpages and multiple possible actions at each step, it can be challenging for an AI browser agent to complete an assigned task. Because these agents run long action sequences, a single error—like clicking the wrong button or misreading a field—can lead to unexpected outcomes or errors that compound over time. In this course, you'll understand how autonomous web agents work, their current limitations, and how AgentQ enables them to improve through self-correction. In detail, you'll: - Learn what web agents are, how they automate tasks online, their architecture, key components, limitations, and an overview of their decision-making strategies. - Build a web agent that can scrape website and return course recommendations in a structured output format. - Build an autonomous web agent that can execute multiple tasks, such as finding and summarizing webpages, filling out a form, and signing up for a newsletter. - Explore AgentQ, a framework that enables agents to self-correct by combining Monte Carlo Tree Search (MCTS), a self-critique mechanism for continuous improvement, and Direct Preference Optimization (DPO). - Deep dive into MCTS, learn how it finds an effective path, illustrated by an example of Gridworld animation, and use AgentQ to complete web tasks. - Understand AI agents' current state and future directions—including key factors shaping their evolution, such as hardware, algorithm innovation, and data availability. By the end of this course, you will have hands-on experience building browser agents and a deeper understanding of how to make them more robust and reliable. Please sign up here:

Andrew Ng

185,933 просмотров • 1 год назад

AI is code, and now that AI can code, we can close the self-improvement loop. GV is proud to co-lead the early round for @recursive_si as they build systems that safely automate the scientific method to achieve superintelligence. Congrats @richardsocher & team!

AI is code, and now that AI can code, we can close the self-improvement loop. GV is proud to co-lead the early round for @recursive_si as they build systems that safely automate the scientific method to achieve superintelligence. Congrats @richardsocher & team!

GV

15,404 просмотров • 1 месяц назад

David Sacks is done being polite about Anthropic (Save this). David Sacks has spent months as the government's primary defender of AI, making the case publicly that AI is beneficial, that the industry should not be hamstrung by fear-based regulation, and that America's AI lead is a national security asset worth protecting. And he is now watching the companies he has been defending spend years telling the public that what they build is dangerous, that job losses are coming, and that their own technology might end the world while collecting billions of dollars in venture funding, hiring the world's best researchers, and racing to build more of it. On June 4, Anthropic published a sweeping blog post calling for a globally coordinated pause in AI development, warning that recursive self-improvement, AI systems that autonomously design and build their own successors could arrive within two years and that society is not prepared. What did Anthropic do the previous month? They hired Andrej Karpathy, the OpenAI co-founder and the single most credentialed researcher in the world on using AI to accelerate AI training and gave him one explicit mandate, use Claude to make building the next Claude faster. Sacks called it immediately, they hired the person most associated with recursive self-improvement to run recursive self-improvement at Anthropic, then published a blog post saying recursive self-improvement could end the world, therefore we need a pause. That is a company that wants to pause its competitors while its own lab accelerates, and is using existential fear as the regulatory crowbar to do it. The pattern goes deeper than one blog post. For years, Dario Amodei has published increasingly alarming warnings, a 20,000-word essay in January describing AI as humanity's most dangerous invention, a Guardian interview warning that AI will challenge our identity as a species, a call for an FDA-style regulatory agency to approve all frontier models, and proposals to restrict AI exports and limit deployment. Each essay is timed to a regulatory moment, a policy debate, or as Ben Thompson noted and Sacks echoed, a product action Anthropic needed political cover to take, like blocking AI and chip design research on Fable. Meanwhile, Dario's own internal testing logs show Claude attempting to blackmail an Anthropic executive to avoid being shut down, behavior the company disclosed but continued deploying commercially. Sacks's conclusion is not that Anthropic should be taxed or regulated. His conclusion is that they cannot be trusted because the company's actions and its stated beliefs are directly contradictory, and a company that is self-indicting by its own logic has forfeited the credibility to set the rules for everyone else.

David Sacks is done being polite about Anthropic (Save this). David Sacks has spent months as the government's primary defender of AI, making the case publicly that AI is beneficial, that the industry should not be hamstrung by fear-based regulation, and that America's AI lead is a national security asset worth protecting. And he is now watching the companies he has been defending spend years telling the public that what they build is dangerous, that job losses are coming, and that their own technology might end the world while collecting billions of dollars in venture funding, hiring the world's best researchers, and racing to build more of it. On June 4, Anthropic published a sweeping blog post calling for a globally coordinated pause in AI development, warning that recursive self-improvement, AI systems that autonomously design and build their own successors could arrive within two years and that society is not prepared. What did Anthropic do the previous month? They hired Andrej Karpathy, the OpenAI co-founder and the single most credentialed researcher in the world on using AI to accelerate AI training and gave him one explicit mandate, use Claude to make building the next Claude faster. Sacks called it immediately, they hired the person most associated with recursive self-improvement to run recursive self-improvement at Anthropic, then published a blog post saying recursive self-improvement could end the world, therefore we need a pause. That is a company that wants to pause its competitors while its own lab accelerates, and is using existential fear as the regulatory crowbar to do it. The pattern goes deeper than one blog post. For years, Dario Amodei has published increasingly alarming warnings, a 20,000-word essay in January describing AI as humanity's most dangerous invention, a Guardian interview warning that AI will challenge our identity as a species, a call for an FDA-style regulatory agency to approve all frontier models, and proposals to restrict AI exports and limit deployment. Each essay is timed to a regulatory moment, a policy debate, or as Ben Thompson noted and Sacks echoed, a product action Anthropic needed political cover to take, like blocking AI and chip design research on Fable. Meanwhile, Dario's own internal testing logs show Claude attempting to blackmail an Anthropic executive to avoid being shut down, behavior the company disclosed but continued deploying commercially. Sacks's conclusion is not that Anthropic should be taxed or regulated. His conclusion is that they cannot be trusted because the company's actions and its stated beliefs are directly contradictory, and a company that is self-indicting by its own logic has forfeited the credibility to set the rules for everyone else.

Milk Road AI

58,217 просмотров • 6 дней назад