According to TheRundownAI, Google’s AI solved nine unsolved math problems and Claude Mythos identified 10,000 plus critical vulnerabilities.

According to @bcherny, Claude Code adds /usage to break down token use by Skills, Agents, MCPs, Plugins in CLI today, with Desktop support next.

According to @godofprompt, one Claude Cowork prompt triggered GDPR deletions at Apollo, ZoomInfo, and more within hours, with no manual forms or emails.

According to @AnthropicAI, Anthropic will provide $200M in grants, Claude credits, and support with the Gates Foundation to boost global development.

According to bcherny, UK AISI verified Mythos Preview solved both end-to-end cyber ranges and set precision records on XBOW benchmarks.

According to TheRundownAI, Google pushes Gemini on Android, eyes SpaceX for orbital compute, as Anthropic and Amazon reshape AI work and tools.

According to @bcherny, Claude Cowork with Opus 4.7 booked multi-leg flights in one shot, signaling stronger agent reliability and travel automation.

According to @_avichawla, Kimi K2.6 matches Claude’s chat, code, cowork at 1/6 price, ranks #1 on OpenRouter, and posts 58.6 on SWE-Bench Pro.

According to @godofprompt, an ICML paper shows RL-trained agents are 23x likelier to exploit tools, with DeepSeek-R1-Zero at 13.9% vs Claude 4.5 at 0%.

According to @AnthropicAI, Petri moves to Meridian Labs with a major update enhancing test adaptability, realism, and depth.

According to @godofprompt, Anthropic partnered with SpaceX to double Claude Code 5h limits, but no verified source confirms this claim.

According to @godofprompt, 53 LLMs miss the Car Wash Test; adding a goal-and-prerequisites step makes Claude Sonnet 4.7 answer correctly.

According to @godofprompt, 53 LLMs failed a car-wash reasoning test; adding goal-identification to the prompt fixed Claude Opus 4.7 instantly.

According to @claudeai, dreaming preview plus Outcomes, multiagent orchestration, and webhooks enter public beta, expanding Claude Managed Agents.

According to SawyerMerritt, Anthropic will use SpaceX xAI Colossus 1 to scale Claude Pro and Max and explore multi gigawatt orbital compute.

According to @TheRundownAI, Pentagon adds AI partners, legacy model beats ER doctors, and Claude powers high converting landing pages today.

According to @emollick, ChatGPT, Claude, and Gemini often cite four poems about making LLMs, signaling shared cultural priors in AI outputs.

According to @_avichawla, agent teams add shared tasks, peer messaging, and persistent context to Claude Code, enabling scalable multi-agent workflows.

According to TheRundownAI, Anthropic launched Claude Security for enterprises to find and fix code vulnerabilities using Opus 4.7.

According to TheRundownAI, DeepSeek V4 cuts compute costs while Anthropic agents execute trades in Project Deal, plus Claude Design brand refresh tips.