OpenAI has shared ten results discovered while evaluating an unreleased model, resolving or making substantial progress on long-standing open problems in mathematics and theoretical computer science. Simultaneously, the Qwen 3.8-Max model with 2.4 trillion parameters is now available, delivering improvements across coding, work, research, and long-horizon tasks.
- OpenAI's evaluation uncovered advances in high-dimensional geometry, coding theory, arithmetic circuit complexity, group theory, operator algebras, quantum complexity, lattice cryptography, and extremal combinatorics.
- Qwen 3.8-Max offers comprehensive improvements for complex tasks with greater reliability, with open weights scheduled for release next week.
- Microsoft is testing MAI Realtime, a bidirectional full-duplex voice model that can listen and speak simultaneously.
These developments highlight significant strides in both theoretical mathematical proof capabilities and practical large language model applications.