The two biggest AI labs just dropped major upgrades on the same day, while a former OpenAI VP is building something entirely different with atoms instead of tokens.
01
Anthropic's AI just got significantly smarter across the board
Anthropic released Claude Sonnet 4.6 with major improvements in coding, computer use, long-form reasoning, agent planning, knowledge work, and design. The upgrade touches every core capability that makes AI useful for real work, from writing code to controlling computers to planning multi-step tasks.
Why it matters: This is the kind of broad capability jump that changes what's possible in your daily workflow — better coding help, smarter document analysis, and AI that can actually follow through on complex requests.
Chinese AI lab MiniMax launches M2 model with new Agent features
MiniMax unveiled their M2 model alongside new Agent capabilities, marking another step forward for Chinese AI companies in the global competition. The release focuses on what they call "ingenious simplicity" in both the underlying model and the agent interface design.
Why it matters: Chinese labs are shipping competitive models at an increasingly rapid pace, giving businesses more options and keeping pressure on Western companies to innovate faster.
Banks are getting AI account managers that actually understand your history
OpenAI showcased Gradient Labs, which uses GPT-4.1 and the new GPT-5.4 mini models to power AI agents that handle banking support workflows. The system promises low latency responses while maintaining the reliability banks need for customer service.
Why it matters: This means your bank might soon have an AI that actually understands your account history instead of making you repeat yourself to three different representatives.
Former ChatGPT co-creator is building AI for the physical world
Liam Fedus, former VP of post-training at OpenAI and co-creator of ChatGPT, appeared on No Priors to discuss his new company Periodic Labs. Instead of building better chatbots, he's using AI to tackle materials science and chemistry — what he calls "an AI foundation lab for atoms." The conversation reveals how top AI talent is moving beyond language models to solve problems in the physical world.
Why it matters: While everyone focuses on chatbots, some of the smartest people in AI are quietly working on using these same techniques to discover new materials and drugs.