Title: AI Competition Intensifies as Models Advance and Access Expands

01GPT-5.6 Solves 30-Year-Old Convex Optimization Problem

OpenAI's latest model, GPT-5.6, has achieved what researchers consider a significant milestone in mathematical research by solving a 30-year-old convex optimization problem. According to a Reddit thread in the mathematics community, GPT-5.6 used a carefully constructed prompt to close a longstanding gap in convex optimization theory. The announcement generated significant discussion, with 305 comments and 477 points on Hacker News, indicating strong interest from both the academic and technical communities.

The breakthrough represents a potential shift in how AI systems might contribute to formal mathematical research. Rather than merely pattern-matching on existing proofs, the model appears to have demonstrated the ability to extend mathematical knowledge into previously unresolved territory. Researchers are now examining whether this approach could be applied to other longstanding open problems in mathematics and theoretical computer science.

02Moonshot AI Releases Kimi K3, Sparking 'Full AI Communism' Debate

Chinese AI company Moonshot AI has released Kimi K3, a new version of its flagship model that has intensified debates about the global AI landscape and development trajectories. The release has prompted TechCrunch to question whether Kimi represents a threat or a different kind of AI development philosophy. The model has drawn particular attention for triggering discussions around the concept of "full AI communism," a term describing a future where AI capabilities are widely shared rather than concentrated.

The timing of Kimi K3's release comes amid heightened competition between US and Chinese AI companies. Analysts note that the model's capabilities appear competitive with leading Western models, raising questions about the global distribution of AI power. The debate has extended beyond technical capabilities to encompass broader questions about AI governance, access, and the implications of rapid capability advancement across geopolitical lines.

03Anthropic Makes Claude Fable 5 Permanent Across Subscription Tiers

Anthropic has reversed its earlier plans to remove Fable 5 from subscription accounts, announcing that the model will become a permanent fixture across its Max and Team plans. According to Simon Willison's coverage, beginning July 20, Claude Fable 5 will be included in all Max and Team Premium plans at 50% of limits, with Pro and Team Standard users retaining access through usage credits plus a one-time $100 credit. The company explicitly acknowledged that competition from GPT-5.6 Sol and Kimi K3 made their previous removal plans untenable.

The reversal signals growing competitive pressure in the AI market as OpenAI and Chinese competitors release capable new models. Anthropic's decision to make Fable 5 permanent across more tiers appears designed to retain and attract users who might otherwise migrate to competing platforms. This pricing shift represents a notable strategic adjustment in the ongoing battle for AI assistant market share.


Also today

ShareXLinkedIn