OpenAI Blog
A contest focused on algorithms that generalize from previous experience has concluded its first round. The organizers did not specify a participant count, prize amount, or performance benchmarks in this announcement. Results from the competition will inform further iterations of the contest format.
The first run of our Retro Contest—exploring the development of algorithms that can generalize from previous experience—is now complete.
AI Agent Wins First Place in Optimization Programming Contest
Sakana AI
Open R1: Update #3
Hugging Face Blog · 1 year ago
What Parameter Golf taught us about AI-assisted research
OpenAI Blog · 2 months ago
Our First Proof submissions
OpenAI Blog · 5 months ago
Quantifying generalization in reinforcement learning
OpenAI Blog · 7 years ago
The daily briefing
TLDRocket reads 60+ sources, removes duplicate coverage, and summarises the day in two minutes. Free, no spam, unsubscribe anytime.
TLDRocket
21 July 2026
The open-weight AI model market hit an inflection point today as Alibaba and Kimi released trillion-parameter competitors that challenge the closed-model orthodoxy—just as Washington moved to restrict them. Alibaba's Qwen 3.8 Max (2.4 trillion parameters) and Kimi K3 (2.8T) are going open-weight, with K3 ranking first on Frontend Web App Arena and fourth on agentic benchmarks, signaling that scale and transparency can coexist. The timing matters: while U.S. policymakers weigh layered restrictions on Chinese open models, technical experts are pushing back hard, arguing that closed gardens actually hurt security and competition more than they help. The real fragmentation risk isn't technical—it's regulatory.
Read the full briefing →