Trends

AICombined annualized revenue for OpenAI and Anthropic reached $105 billion by August 2026, up 3.5 times from $30 billion at the start of the year.

Why it matters: The piece tests whether OpenAI and Anthropic's hypergrowth reflects a temporary coding-agent spike or durable progress, using revenue scale to frame the question.

AIFireworks AI reports that DeepSeek V4 Pro 0813 scored 95.2% on SWE-Bench Verified, ahead of Kimi K3 at 92.6% and Fable 5 at 85.4%.

Why it matters: The post reports a model's benchmark results and cost per solved task against Fable 5, and it notes the Java gap plus the limits of routing as a ceiling, not a deployable policy.

AIMark Chen, who is affiliated with OpenAI, said the company signed on for more than 4 GW of capacity with NVIDIA. He called this the scale that frontier training demands.

Why it matters: The post announces a compute deal of more than 4 GW with NVIDIA, a scale figure that shows how large frontier training infrastructure has become.

AIQuoc Le, Jeff Dean, Sanjay Ghemawat, and Oriol Vinyals announced they are co-founding Discovery Loop, a Public Benefit Corporation.

Why it matters: The announcement names the co-founders and a public benefit structure, showing how this new company frames automating machine learning, science, and engineering.

AIMiniMax has upgraded its Agent, renamed Mavis, and introduced Agent Teams that run multiple role-based Agents in parallel on desktop.

Why it matters: The post explains why multi-Agent helps long tasks and where its verification, token, and aggregation costs come from, useful for judging when a team setup beats a single Agent.

AIOpenAI has shipped GPT-5.5 Instant to ChatGPT, rolling out to everyone over the next couple of days.

Why it matters: The quoted post gives concrete targets for the update, factuality and hallucination reduction, useful for judging whether the default ChatGPT model changed in practice.

AIResearchers trained a humanoid with 22-DoF dexterous hands mainly on over 20,000 hours of egocentric human video, with no robot in the loop, to perform tasks such as assembling model cars and folding shirts.

Why it matters: The source reports a measured scaling law linking human video volume to robot action loss, which bears on how dexterous robot training data might be collected and reused.

AIAndrej Karpathy's year-in-review lists the LLM paradigm changes he found most notable in 2025.

Why it matters: Karpathy ties the year's shifts to RLVR, jagged capability, and local agents, giving readers a framework for judging how LLM progress is changing.

AIARC Prize reports that the top Kaggle entry reached 24% on the ARC-AGI-2 private dataset at $0.20 per task, and that all winning solutions and papers are open source.

Why it matters: The post links 2025 competition results to a broader argument about refinement loops, showing how benchmark outcomes are being read as evidence of AI reasoning progress.