Company / organization
OpenAI
By and about OpenAI · page 7
Blog Post: The Second Half by Shunyu Yao
OpenAI researcher Shunyu Yao argues that AI's second half will shift from developing new training methods to defining the right problems and evaluation metrics, with pre-training priors mattering more than algorithms.
newsgpt-oss-20b (free)
OpenAI released gpt-oss-20b, a 21 billion parameter open-weight model under the Apache 2.0 license.
newsQwen3 Coder 30B A3B Instruct
Alibaba released Qwen3 Coder 30B A3B Instruct on July 31, 2025, a 30.5B parameter mixture-of-experts model with 256K token context designed for code generation, repository understanding, and tool use.
newsWindsurf's Updates: Google asset deal and Cognition acquisition
Google acquired Windsurf's assets while Cognition Labs acquired the company itself, allowing employees to join Cognition after the deal structure left them out of Google's asset purchase.
newsGoogle Keeps on Shipping: Gemini 2.5 Pro and Veo3
Google launched Gemini 2.5 Pro, now second only to OpenAI's o3-pro, and Veo3, ranked #2 in video generation models, as the company continues shipping new AI products.
newsOpenAI Leads with o3-pro at 80% Lower Cost
OpenAI's o3-pro matches o3 performance on ARC1 and ARC2 benchmarks while cutting costs by 80%, pricing it near 4o-mini levels.
newsMeta Takes 49% Stake in Scale AI for $14.3 Billion
Meta acquires 49% stake in Scale AI for $14.3 billion and appoints founder Alexandr Wang as head of AI, though critics question his operational abilities despite fundraising success.
newsLearning: Best AI Guides and YouTube Channels
OpenAI, Google, and Anthropic have released official AI learning guides, plus a curated list of the top 10 YouTube channels for beginners.
newsOpenAI Codex: cloud SWE agent built on codex-1 reasoning model
OpenAI launches Codex, a cloud-based software engineering agent built on its o3 reasoning model that executes coding tasks like feature writing, bug fixes, and tests in parallel within sandboxed environments, completing work in 1–30 minutes with…
newsOpenAI acquires Windsurf for $3B
OpenAI acquires Windsurf for $3 billion, consolidating control over AI application layers and reducing reliance on competing models.
newsOpenAI publishes a guide on when to use which model
OpenAI published a guide explaining which model to use for different tasks, covering performance, cost, and capability tradeoffs.
newsGPT model stopped learning Croatian due to downvoting users
OpenAI's GPT model stopped improving on Croatian because users in that language community downvoted responses more frequently, reducing the feedback signal for learning.
newsOpenAI releases o3 and o4-mini reasoning models
OpenAI released o3 and o3-mini reasoning models, with o3 scoring 87.7% on GPQA Diamond and 71.7% on SWE-Bench Verified, while o3-mini offers lower-cost performance for coding tasks.
newsOpenAI drops GPT-4.1 with Mini and Nano variants
OpenAI released GPT-4.1 with Mini and Nano variants, offering a 1M token context window, 55% score on SWE-Bench Verified coding tasks, and pricing from $0.40/$1.60 per 1M tokens for the smaller models.
newsFun Fact: "please and thank you" are costing OpenAI millions
https://x.com/sama/status/1912646035979239430 If you're curious to know how the sausage is made 🥩 + 🥖 = 🌭 1\. Bookmark tweets, videos and blog posts over a month 2\. Combine them into a sheet https
newsAnthropic Claude Sonnet 3.7 (Aurora) 🛸
Anthropic released Claude Sonnet 3.7 (Aurora) on March 10, 2025, achieving a ChatEval score of 0.85 and an XGLUE score of 0.88, with the company also introducing prompt caching to reduce inference costs.
newsOpenAI releases GPT-4.5 (Orion) 🚀
OpenAI launched GPT-4.5 (codenamed Orion) on February 27, 2025, emphasizing natural conversation and reduced hallucinations with improved factual accuracy across benchmarks, though it lags specialized models in complex reasoning.
newsManus: DeepResearch + Operator agent from China
Chinese startup Butterfly Effect released Manus, an autonomous AI agent powered by Claude 3.5 and Qwen that performs real-world tasks like building websites and analyzing stocks, outperforming OpenAI's Deep Research on the GAIA benchmark.
newsOpenAI Launches O3, Operator, and DeepResearch
OpenAI launches O3 (scoring 87% on ARC), Operator (an AI agent for web browsing and clicking), and DeepResearch (multi-source research reports) as part of a new $200/month Pro subscription.
newsAnthropic Is Eating OpenAI's Market Share
Anthropic is eating OpenAI's market share https://x.com/itsandrewgao/status/1885144792323285183