Invested by Aleph · 21 October 2026
At the source: Michael Eisenberg ↗ · audio
Josh Kopelman, founding partner of First Round Capital and Half.com founder, on OpenAI and Anthropic's valuations and whether AI breaks seed investing. On this episode of Invested, Michael Eisenberg talks with Josh Kopelman about building Half.com at the height of the dot-com era, when the company convinced Halfway, Oregon to rename itself Half.com, put coupon codes on 2 million fortune cookies, and sold to eBay in 2000. Josh compares that bubble to today's AI boom and explains why First Round's fund grew from a little under $150 million in 2004 to about half a billion without changing its strategy. He also argues that late-stage bets on SpaceX, Anthropic and OpenAI are a cyclical trade, while seed is a business that can make money in both risk-on and risk-off markets. Josh goes back to Infonautics, the company he co-founded as a Wharton student, which scanned newspapers and magazines into Homework Helper for AOL, CompuServe and Prodigy before search engines existed. He explains why founders are cartographers, why he stays up until 3 a.m. building MCP servers with Claude Code even though he is not a strong engineer, why he doesn't think AI breaks the seed model, and how The Philadelphia Inquirer, where he served as chairman, went from "default dead" to "default alive." Josh also describes publicly calling out Web Summit founder Paddy Cosgrave's tweets after October 7th, what it cost him, and what First Round looks for in Israeli founders.
As published with the episode at Invested by Aleph ↗
No summary of this episode yet.