Billionaire tech investor says 'valuations in AI are at a bubble' (www.cnbc.com)

Thoma Bravo co-founder Orlando Bravo warned that valuations for AI companies are "at a bubble," likening the stretch to the dot‑com era while noting a crucial difference: large corporations with healthy balance sheets are now financing much of the activity. Bravo — whose private equity firm manages over $181 billion and specializes in enterprise tech and cybersecurity — argued buyers can’t rationally price a company doing $50 million in ARR at $10 billion without it ultimately producing roughly $1 billion in free cash flow to satisfy investors. He highlighted extreme examples across public and private markets: OpenAI’s recent secondary implied valuation near $500 billion (with projected $13 billion revenue in 2025), Nvidia’s plan to invest up to $100 billion to lease chips and build supercomputing capacity, Palantir and AppLovin’s market caps in the hundreds of billions, and seed-stage firms like Thinking Machines Lab reaching multibillion-dollar valuations. The practical implication for the AI/ML community is twofold: there’s elevated execution and managerial risk as startups are given sky‑high price tags that assume near‑perfect scaling, yet corporate balance‑sheet support could prevent a sharp systemic collapse typical of past bubbles. For investors and builders this means greater scrutiny of unit economics, ARR-to‑FCF conversion pathways, capital intensity (especially for AI infrastructure), and consolidation dynamics — areas where technical progress, cost of compute, and demonstrable enterprise adoption will determine which valuations are ultimately justified.