Anthropic dominance/restrictions fueling OSS agent surge, Meta comms push
Key Questions
What is Anthropic's reported annual recurring revenue and compute scale?
Anthropic has reached $30B ARR with 3.5GW of compute. They previewed Claude Mythos for interpretability research.
What major contract did Meta sign with Anthropic?
Meta signed a contract for 60T Claude tokens, reportedly Anthropic's largest and possibly the biggest in software history. This supports Meta's AI initiatives.
What acquisition did Meta make for AI agent communications?
Meta acquired Moltbook for $2B, a Reddit-style forum where AI agents post and interact, including inventing concepts like 'Crustafarianism'. It enhances OpenClaw agent comms.
How have Claude restrictions affected open-source model usage?
Claude bans and paygates on OpenClaw have spiked usage of OSS models like GLM-5.1 and GPT-5.4, with GPT-5.4 up 8.9% in a week. This fuels the OSS agent surge.
What is the popularity of OpenClaw?
OpenClaw has 247k stars on GitHub. Variants include NemoClaw, Hermes, and Genspark Claw.
What are the risks associated with n8n MCP?
n8n's Model Context Protocol (MCP) carries risks, mentioned alongside the OSS agent surge.
What is Anthropic's Project Glasswing?
Anthropic's Project Glasswing is deemed necessary, linked to their interpretability efforts like Claude Mythos Preview.
How has Anthropic expanded its compute resources?
Anthropic upped compute deals with Google and Broadcom amid skyrocketing demand. This supports their growth to 3.5GW.
Anthropic $30B ARR/3.5GW compute/Mythos preview interpretability; Meta 60T Claude tokens contract + $2B Moltbook acq for OpenClaw agent comms; Claude bans/paygates spike OSS/GLM-5.1/GPT usage; OpenClaw 247k stars/NemoClaw/Hermes/Genspark Claw; n8n MCP risks.