Dear VCs, Just Give Up
Summary
Michael Dempsey satirically tells boutique seed VCs to quit: mega-funds are crowding seed, only a handful of companies matter, founder sourcing is fully mapped, and AI-only herding plus two-year AGI timelines make a ten-year fund pointless.
Mega-funds win seed by brand, not skill.
A $5B fund will fight for 10-15% of a two-person company; founders pick the known name over a boutique that may not exist in a decade.
Fewer companies are treated as consequential.
Accessible intelligence concentrates outcomes; consensus says if you are not in OpenAI, Anthropic, Anduril, or a neolab, you do not matter.
Legibility is sold as a solved sourcing problem.
Elite schools, contests, and spinouts, plus CRM-mapped GitHub, PhD, and dropout pipelines, make the founder universe look fully known.
All-in AI plus AGI timelines are a herd instruction.
Putting the entire fund in AI, skipping every industry it will remake, and abandoning a ten-year fund because AGI is "2-3 years away" is the consensus he is mocking.
The closing inversion is the thesis.
"Incumbents destroy upstarts" is the opposite of venture history; giving up, rather than doing something slightly different, is what the essay forbids.
Technology has taught us one thing over and over; that the big incumbents destroy the upstarts.
Read the original ↗Save to your Sponge
Summarized with Sponge. Quotes are the author’s own words.