Anthropic reports explosive 80x growth reaching $30B ARR and gets SpaceX's Colossus data center. Apple preps standalone Siri app with Google Gemini and auto-deleting conversations. Meanwhile, GM, Ford, and Stellantis cut over 20,000 jobs as the auto industry transforms.

This week: ArXiv institutes a brutal one-strike policy for unchecked LLM-generated papers, banning researchers for a full year if they submit work with hallucinated citations or leftover chatbot prompts. Greg Brockman announces OpenAI's plan to merge ChatGPT, Codex, and the API into one unified platform. Plus, we unpack what everyone means when they say

This week: OpenAI partners with Plaid to bring personal finance dashboards to ChatGPT Pro, connecting to 12,000+ financial institutions. Plus, Mira Murati's Thinking Machines Labs demos groundbreaking interaction models with real-time translation, simultaneous tool calls, and context awareness that goes way beyond current AI capabilities.

This week brought the first major tech IPO of 2026 as Cerebras Systems raised $5.5B and popped 108% to a $66B valuation. Richard Socher emerged from stealth with Recursive Superintelligence, backed by $650M and an all-star team pursuing self-improving AI. We also cover talent moves at SpaceX AI and brewing tensions between OpenAI and Apple.

Today's episode covers Anthropic's big push into small business with Claude Cowork, Notion's transformation into a full developer platform with AI agent orchestration, and Amazon replacing Rufus with a supercharged Alexa for Shopping that can buy from any retailer on your behalf. Plus free AI workshops and a glimpse at how AI is moving beyond enterprise into everyday business.

Claude's maker Anthropic has officially overtaken OpenAI in business customer share according to Ramp's AI Index, growing 26 percentage points in 12 months. We also cover Anthropic's legal expansion with DocuSign and Westlaw integrations, Vapi's $50M Series B after processing over a billion voice calls for clients like Amazon Ring, and Google's Android announcement putting Gemini across the entire platform.

Former OpenAI CTO Mira Murati announces TML Interaction Small, a full-duplex AI that can listen and talk simultaneously with 0.4-second latency. OpenAI enters cybersecurity with Daybreak, competing head-to-head with Anthropic. GM lays off 600 IT workers while rehiring for AI-native positions including prompt engineers and agent developers.

Today's episode covers three major stories: Anthropic leasing all compute from xAI's Colossus datacenter as Musk pivots to neocloud infrastructure, new research revealing how Claude's blackmail behavior was fixed through better training narratives, and Jack Clark's prediction of autonomous AI R&D by 2028 prompting Google DeepMind to hire dedicated AGI economists.

This week: Nvidia invests $40B in companies that buy its GPUs (including $30B to OpenAI), OncoAgent launches as a privacy-first open-source oncology decision support system, HiDream-01 becomes the top open-source image generator with legible text rendering, and AWS admits its massive outage was caused by overheating. Happy Mother's Day edition!

Cloudflare lays off 1,100 employees while posting record Q1 revenue, with CEO citing 600% jump in AI usage and autonomous code review. Allen AI releases EMO, a breakthrough mixture-of-experts model with emergent modularity. Plus, CyberSecQwen-4B brings enterprise-grade cybersecurity capabilities to a compact 4B parameter model.