OpenAI's Math Breakthrough, Meta's Muse Agent, and Industry Funding
OpenAI reports a major math breakthrough using 10,000 agents, Meta launches its personal AI agent Muse, and Cognition secures $2B in funding. Additionally, new developments from DeepMind, Google Cloud, and Qualcomm are reshaping the AI landscape.
Podcast В· 3 min
OpenAI Claims Solution to Millennium Prize Math Problem
OpenAI has reported that an internal model, described as significantly more capable than the recently released GPT-6 Astra, has successfully solved the Navier-Stokes Millennium Prize Problem. The system utilized a swarm of 10,000 agents working in parallel for 88 hours to generate the proof, which was then formalized and verified using Lean. The announcement has sparked controversy, with mathematicians Tristan Buckmaster and Levent Alpöge alleging that OpenAI may have leveraged their prior work. OpenAI has denied accessing their specific research, stating the proof was independently produced.
DeepMind Releases AlphaGenome Atlas
Google DeepMind has launched the AlphaGenome Atlas, a predictive map that identifies the molecular effects of all 9 billion possible single-letter DNA mutations in the human genome. This resource is intended to assist researchers in understanding the biological impact of genetic variations.
Meta Introduces Muse Personal AI Agent
Meta has launched Muse, an always-on personal AI agent designed to handle autonomous tasks such as booking reservations, managing emails, and shopping. The agent operates within its own cloud virtual machine, allowing it to navigate websites and fill out forms. Muse integrates with services like Gmail, Spotify, and OpenTable, and is capable of coding its own integrations where necessary. Currently available to U.S. users, the service offers a free tier alongside $20 and $100 monthly subscription plans.
OpenAI Rolls Out ChatGPT Images 2.5
OpenAI has released ChatGPT Images 2.5, an updated image generation model that reportedly reduces generation time by up to 50% compared to its predecessor. The update includes improved editing capabilities, allowing for more precise modifications, as well as new features such as sketch-to-image tools, templates, and shared image functionality.
Inception Labs Releases Mercury 2.5
Inception Labs has introduced Mercury 2.5, a new diffusion model designed for speed. The model generates text in parallel rather than sequentially, claiming to achieve high-quality performance at speeds exceeding 1,100 tokens per second.
Cognition Raises $2B at $48B Valuation
Cognition, the developer behind the AI software engineering platform Devin, has secured over $2 billion in a new funding round, bringing its valuation to $48 billion. The company continues to focus on scaling its software engineering capabilities.
Qualcomm, Amazon, and Accenture Announce AI Partnerships
Several major industry partnerships were announced. Qualcomm and Amazon have entered a multi-generation deal to develop custom AI inference chips and high-speed connectivity for AWS data centers. Additionally, Google Cloud and Accenture have formed a new Gemini Enterprise unit, which aims to deploy 1,000 forward-deployed AI engineers directly within client organizations to accelerate AI adoption.