X：Rohan Paul

Jensen Huang on "distillation" In this interview with axios, he was asked this question: "Should open source model companies be allowed to distill closed models" "Distillation, learning from AI, learning from other people, and learning from other sources of knowledge, is fundamental to intelligence. We are constantly learning from other people. I am learning from you through the questions you are asking, and you are learning from me. All day long, we are learning from one another. AI also has to learn from something. The original AI models, whether they were open or closed, were trained on previously created knowledge from the internet. Now, AI is generating more content than humans. In a few more years, the internet could be 99% AI-generated content, and that content will have been created by some form of AI. As a result, AI systems will constantly be distilling knowledge and intelligence from other AI systems. The fact that AI can learn is a good thing. We want AI systems to be intelligent because a smarter AI can also be a safer AI." ---- From "Axios" YouTube channel, (full video link in comment)

In a new allegation, the U.S. government has accused 6 Chinese AI firms of using large-scale distillation to copy American model capabilities. The NSA, CISA and FBI published joint advisory AA26-251A, naming DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun and Z .AI as running industrial-scale distillation campaigns against U.S. frontier models since at least late 2024. The agencies say requests moved through a gray market of API proxies called transfer stations, bulk premium subscriptions shared across developer teams, and third-party aggregators that strip account metadata. Moonshot AI allegedly distilled 17 U.S. models, including Anthropic's Claude Fable 5, to train Kimi-K3. They claim DeepSeek's prompts pushed models to write out hidden chain-of-thought steps, which transfers reasoning method rather than finished answers. So they say that DeepSeek's headline training cost only covers the compute it burned. The advisory argues that number is misleading because it leaves out what the training data actually cost, which DeepSeek allegedly obtained by distilling U.S. models rather than producing it through its own research. So the cheap-training story rests on data someone else paid to create. The mitigation section asks U.S. labs to serve suspected distillers subtly degraded responses without telling them. Its detection indicators are behavioral, covering sustained 24/7 usage, new accounts at immediate maximum throughput, and traffic optimized for cache hits. Those patterns also describe an ordinary enterprise agent fleet, leaving each provider to decide which customers receive an undisclosed downgrade.

Demand for Astra is really unprecedented. We're pulling all the levers possible to sustain the demand, but I've not seen anything like it until now and we went through very steep growth before. Priority will always be to keep excellent service for existing users, but we might have to pause new Pro subscriptions for a bit if this continues.

Mark Zuckerberg talks about how he is using Meta’s newly released Muse agent. something much closer to a persistent personal worker than a chatbot. helping his 3-year-old daughter likes baking, monitor when mountain climbing permits become available and secure them for their weekend trips, and also for physical training.. He installed cameras in his MMA gym and has Muse watch the footage and give him feedback on his performance. Zuckerberg gives it ongoing goals and projects, and the agent can continue working on them, remember what happened, and proactively suggest what it should do next. ---- From "Sources Podcast and @alexeheath YouTube channel, (full video link in comment)

Meta just released Muse AI assistant for personal tasks, powered by Muse Spark 1.3, Meta's latest model for longer-horizon agentic work - it can connect to email, calendars, shopping, payments and other services, then keep working after the app closes. - each user gets a Muse Secure VM, with a separate Sentinel agent that checks outbound data and actions before they reach the network. - Meta is making Muse free to use for up to 100M tokens per week, with subscription plans for people who use more compute - Sensitive actions such as purchases or emails require user approval, while credentials stay in Secure Credential Storage so Muse cannot read passwords or payment details. - the model, Muse harness, deterministic code and classifiers screen prompt injections before hostile content can enter the model's context. - currently it launches for U.S. users on web, iOS, Android and WhatsApp, with free access plus $20 and $100 monthly tiers for heavier usage. - the security design moves the trust boundary beyond model behavior into VM isolation, credential separation and a kernel-enforced gate on network actions.

Mark Zuckerberg talks about how he is using Meta’s newly released Muse agent. something much closer to a persistent personal worker than a chatbot. helping his 3-year-old daughter likes baking, monitor when mountain climbing permits become available and secure them for their weekend trips, and also for physical training.. He installed cameras in his MMA gym and has Muse watch the footage and give him feedback on his performance. Zuckerberg gives it ongoing goals and projects, and the agent can continue working on them, remember what happened, and proactively suggest what it should do next. ---- From "Sources Podcast and @alexeheath YouTube channel, (full video link in comment)

full video https://t.co/QqMXZUrhFz

Meta just released Muse AI assistant for personal tasks, powered by Muse Spark 1.3, Meta's latest model for longer-horizon agentic work - it can connect to email, calendars, shopping, payments and other services, then keep working after the app closes. - each user gets a Muse Secure VM, with a separate Sentinel agent that checks outbound data and actions before they reach the network. - Meta is making Muse free to use for up to 100M tokens per week, with subscription plans for people who use more compute - Sensitive actions such as purchases or emails require user approval, while credentials stay in Secure Credential Storage so Muse cannot read passwords or payment details. - the model, Muse harness, deterministic code and classifiers screen prompt injections before hostile content can enter the model's context. - currently it launches for U.S. users on web, iOS, Android and WhatsApp, with free access plus $20 and $100 monthly tiers for heavier usage. - the security design moves the trust boundary beyond model behavior into VM isolation, credential separation and a kernel-enforced gate on network actions.

Introducing Muse, the personal agent that understands your goals and works 24/7 to get things done for you.

OpenAI's unreleased internal model sits on a much higher capability curve than GPT-6 Astra. the one that just solved the 90-year old Navier-Stokes Millennium Prize math Problem. the new model also appears to use additional reasoning compute much more productively. That's how throwing 10,000 agents and enormous compute at Navier-Stokes could produce something qualitatively different from simply running Astra longer.

MASSIVE: OpenAI's 10,000 AI agents, powered by a model significantly more capable than GPT-6 Astra, solved a math problem unresolved for roughly 90 years. Looks like, in this new world, the scaling law may be agent count, not model size: enough capable agents can search far more scientific ideas than any human group ever could. when tens of thousands of strong agents can explore, criticize, and combine ideas, many problems that are "too hard" today may become mostly a compute problem. For this one, for 90 years, mathematicians could not answer whether smooth 3D fluid flow can suddenly break down. OpenAI's proposed proof says it can. To find this solution, OpenAI split thousands of agents into groups exploring different ideas, then shared the best findings between them. The search took 88 hours and used about 130B output tokens across 2.7M messages. GPT-6 Astra then spent another 17 hours checking and converting the proof into Lean, a system that verifies mathematical proofs. That Lean version is public, so mathematicians can inspect the argument themselves.

This model represents a step-function improvement on many benchmarks, and its training is ongoing. Our internal model group arrived at the Navier–Stokes solution in 88 hours, using around 10,000 coordinating AI agents. Throughout the effort, we maintained the strict safeguards—including monitoring and isolation—that we apply to all our frontier evaluations.

We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.