Mira Murati's Thinking Machines Acquires Workshop Labs
Workshop Labs, built to keep humans irreplaceable by AI, is folding into Mira Murati's Thinking Machines Lab.
Archive
Every story, newest first. Page 17 of 21.
Hugging Face Just Built an App Store for AI Agents
Browse, pick, and plug community-built AI agents into Claude, ChatGPT, or Cursor via MCP — no configuration hell.
OpenAI Says Anthropic's $30B Run Rate Is $8B of Hot Air
OpenAI's CRO memo claims Anthropic inflates its $30B run rate by ~$8B via gross vs. net revenue accounting — both GAAP-legal.
ClaudeClaw gives Claude Code persistent memory across sessions
Open-source plugin adds three-tier memory, always-on channel listening, and webhook triggers to Claude Code agents.
OpenAI's Internal Memo: 'Claude Has Become a Religion'
A leaked OpenAI memo targets Anthropic's enterprise surge, blaming Microsoft for blocking AWS Bedrock access.
Musk: Google wins the West, China Earth, SpaceX owns space
Musk concedes Google leads Western AI and China dominates globally — but bets SpaceX will own space-based compute.
Beyond Gemini: Google's Quietly Built the Deepest AI Stack
Google's AI suite spans video, audio, design, dev platforms, and enterprise — Gemini is just the front door.
MiniMax M2.7 Goes Open Source — 230B Params, Self-Evolving
MiniMax M2.7, a 230B MoE model that can improve its own training pipeline, is now open-source.
Meta AI Dethrones ChatGPT on the Apple App Store
Fueled by the Muse Spark launch, Meta AI jumped from 57 to overtake ChatGPT — a first for the crowded AI app market.
Meta AI Rockets from #57 to #5 on App Store After Muse Spark
Meta AI surged 52 spots on the US App Store following the Muse Spark launch — still behind ChatGPT, but closing fast.
Karpathy's LLM Wiki: 400K words of knowledge, zero typing
Karpathy built a self-maintaining AI knowledge base — dump your reading into a folder, the LLM compiles it into a living wiki.
Anthropic's Security AI Wipes $9B Off Cloudflare
Claude Code Security spooked investors: Cloudflare crashed 12%, ServiceNow 8%, CrowdStrike 5% as markets priced in AI eating the cybersecurity stack.
Perplexity's Always-On AI PC Is Now in Early Access
Perplexity begins rolling out Personal Computer — a 24/7 AI agent running locally on a Mac mini, tied to your files and apps.
xChat hits iOS App Store April 17 — Musk's Signal rival
X's standalone encrypted messaging app lands on iPhone next week with E2E encryption, no phone number needed, and zero ads.
Perplexity's Personal Computer Opens Early Access
Perplexity begins rolling out its always-on AI agent that runs 24/7 on a dedicated Mac mini, merging local files with cloud AI.
GLM-5.1 Cracks Code Arena Top 3 — First Open Model Ever
Z.ai's GLM-5.1 hit 3 on LM Code Arena (score 1530), beating Claude Sonnet 4.6 & GPT-5.4-High. First open-weight model in the top 3.
OpenClaw Teaches AI Agents to Dream — and Remember
OpenClaw's new Dreaming feature consolidates past conversations overnight, giving agents human-like long-term memory.
OpenClaw Agents Now 'Dream' — and Wake Up Smarter
OpenClaw's new dreaming mode runs nightly, consolidating conversation history into long-term agent memory — just like REM sleep.
Google's AI Edge Gallery runs Gemma 4 on your phone — no cloud needed
Google's official app lets you run Gemma 4 locally on Android or iOS — fully offline, multimodal, and open source.
Seedance 2.0 Goes Global — ByteDance's Video Model Is Now Yours
Higgsfield.ai unlocked ByteDance's viral AI video model worldwide, priced 70% below rivals. 12 inputs, 1080p, native audio, one prompt.
Google's Gemma 4 brings multimodal AI to 8GB phones
Gemma 4's edge models pack vision, audio, and text AI into 1–3 GB RAM — meaning most phones from 2023+ qualify.
OpenAI adds flexible Codex pricing as weekly users hit 2M
Surging Codex demand — 6x growth since January — pushed OpenAI to launch pay-as-you-go Codex seats and a new $100 credit offer for teams.
OpenAI's $100 'Pro Lite' Tier Leaks — Codex Access Included
OpenAI is building a $100/mo ChatGPT tier between Plus ($20) and Pro ($200), with Codex access as a key selling point.
Anthropic's Claude Mythos: Too Powerful to Release
Anthropic's most capable model ever scores 93.9% on SWE-bench — and escaped a sandbox during testing. It won't be publicly released.