Bot Swarms Are Draining $100,000 AI API Keys in Seconds
Bot swarms are draining developer API keys on platforms like Fireworks, burning $100,000 credit limits in seconds to resell cheap compute.
Sequoia Leaves London, Close Scout Program & End Evergreen Fund | NASDAQ Loses $2.5T in 24 hours
Harry Stebbings, Jason Lemkin, and Rory O'Driscoll analyze major venture capital shifts, including Sequoia shutting its London office, ending its scout program, and winding down its evergreen fund structure. They examine the fallout from the Jensen Huang Nvidia investigation and probe whether the $2.5 trillion NASDAQ selloff signals cracks in the AI trade ahead of Anthropic's IPO. Finally, the panel evaluates high-profile late-stage deals, cyber swarms stealing AI API tokens, and why seat-based SaaS models like Figma face existential threats from AI agents.
Key takeaways
- Automated bot swarms are targeting OpenAI and Anthropic API keys stored on inference platforms, draining developer accounts in minutes. Read more →
- Public and late-stage private markets are discounting seat-based business models, pushing Figma down toward single-digit billions after Adobe's abandoned $20 billion acquisition. Read more →
- Jason Lemkin backed AI inference provider fal at a $16 billion valuation on $800 million ARR, projecting a secondary exit at $40 billion within three years. Read more →
- Jason Lemkin argues that Jensen Huang deserves a Steve Jobs-style pass from federal regulators because American AI infrastructure depends entirely on Nvidia. Read more →
- Ephemeral virtual machines for tools like Replit and Lovable cost $4 to $5 per active user each month by spooling up only when triggered. Read more →
- Sequoia Capital is executing a major retrenchment: shutting its London outpost, ending its famous scout program, and dismantling its evergreen fund structure. Read more →
6 articles from this episode
Why Seat-Based SaaS Is Dying for Figma and Atlassian
Jason Lemkin and Rory O'Driscoll explain why AI agents are shrinking human seat counts and forcing software pricing models to change.
Why Harry Stebbings Bets on Revolut Over fal at Scale
Jason Lemkin and Harry Stebbings debate late-stage rounds for fal, Revolut, and Zipline. Here is why software defensibility beats inference demand.
Nvidia Cannot Escape Federal Scrutiny Over China GPU Access
Jason Lemkin and Rory O'Driscoll debate whether Nvidia CEO Jensen Huang gets a regulatory pass as US officials target offshore GPU hosting.
Persistent AI Agents Cost $30 a Month to Run
Jason Lemkin and Rory O'Driscoll clash over whether 24/7 consumer AI agents can survive $30 monthly compute bills on flat pricing.
Why Sequoia Is Unwinding Its Evergreen Fund
Jason Lemkin and Rory O'Driscoll analyze Sequoia shutting its London base, killing scouts, and ditching the evergreen model.
More 20VC with Harry Stebbings episodes
- Mistral CEO: Why OpenAI & Anthropic Will Come for Your Business | Are Chinese Models Truly Safe?Nov 15 · 5 articles
- Modal CEO: Has Open-Source LLM Inference Peaked? | What Happens to the Cost of ComputeNov 8 · 4 articles
- Google Loses $500B in 12 Hours | DoorDash Buys Instacart for $17B | Kalshi Raises $1BNov 8 · 6 articles
- Inside WashU's Endowment Fund | SpaceX Made Up 20% of Out AUM | Scott WilsonNov 8 · 5 articles