AI for Science: AlphaFold, Materials, Weather, Math — 2026-09-13
OpenAI has claimed a breakthrough solution to the Navier-Stokes Millennium Prize problem, sparking immediate controversy over credit and verification with mathematician Tristan Buckmaster. Simultaneously, Google DeepMind updated its WeatherNext 3 model to integrate raw satellite data, while Chinese media highlighted GPT-6 Astra's superior performance in antibody prediction benchmarks.
AI for Science: AlphaFold, Materials, Weather, Math — 2026-09-13
Top developments
OpenAI Claims Navier-Stokes Solution, Sparking Verification Crisis
On September 8, 2026, OpenAI announced it had solved the Navier-Stokes existence and smoothness Millennium Prize problem, a major milestone in applying AI to formal mathematics. The announcement triggered an immediate dispute involving mathematician Tristan Buckmaster, who reported being "crushed" between AI giants after his own work on the problem was preempted by the model's output. This event marks a pivotal shift where AI systems are not just assisting but potentially outpacing human researchers in solving foundational mathematical puzzles.

Google WeatherNext 3 Integrates Raw Satellite Data
Google DeepMind released an update to its WeatherNext 3 AI weather model on September 9, 2026, which now incorporates raw satellite data alongside traditional inputs to improve forecast accuracy. This enhancement allows the model to better capture high-resolution atmospheric details, continuing the trend of AI models outperforming traditional numerical weather prediction (NWP) methods in medium-range forecasts. The update aims to refine predictions for extreme weather events, building on previous versions that already demonstrated superior skill scores against ECMWF baselines.

GPT-6 Astra Tops Antibody Prediction Benchmarks
Chinese tech media reported that OpenAI’s GPT-6 Astra model achieved top rankings in the DDD Benchmark for antibody prediction, surpassing specialized tools like AlphaFold in certain large-molecule tasks. The coverage highlights how general-purpose large language models are increasingly encroaching on domain-specific scientific AI niches, particularly in drug discovery and structural biology. This development underscores the rapid convergence of general AI capabilities with specialized scientific applications, challenging the dominance of purpose-built models like AlphaFold in specific sub-domains.

Local view
Chinese media outlets such as Phoenix Tech (ifeng.com) are actively analyzing the implications of OpenAI's recent scientific claims, focusing heavily on the competitive dynamics between general-purpose LLMs and specialized scientific models. The coverage emphasizes the "DDD Benchmark" results as evidence that generalist models are becoming viable competitors in hard science domains like antibody design. Meanwhile, Sohu published detailed technical interpretations of the Navier-Stokes proof, exploring its potential applications in fluid dynamics engineering and climate modeling.
Context & numbers
- Weather Model Skill: Recent benchmarks indicate leading AI models (including EPT-2, Microsoft Aurora, and ECMWF AIFS) now outperform traditional NWP on most standard verification metrics at 5–10 day lead times.
- Mathematical Impact: The Navier-Stokes problem is one of seven Millennium Prize Problems established by the Clay Mathematics Institute, carrying a $1 million prize for a verified solution.
On the radar
- Verification of Navier-Stokes Proof: The mathematical community is currently scrutinizing OpenAI's claim; independent verification by human experts or formal proof assistants like Lean is expected in the coming weeks.
- WeatherNext 3 Deployment: Google plans to integrate the updated WeatherNext 3 forecasts into Search, Maps, and Gemini services, with widespread user-facing changes anticipated later this month.
This content was collected, curated, and summarized entirely by AI — including how and what to gather. It may contain inaccuracies. Crew does not guarantee the accuracy of any information presented here. Always verify facts on your own before acting on them. Crew assumes no legal liability for any consequences arising from reliance on this content.