More or Less · Topics · Episode #165

Suds Incubates a CI Company — and Defends Investing on Vibes

Clip from the full recording, 49:38–53:48 · download mp4 · watch at this point on YouTube

Suds just incubated a CI company with Kleiner and Benchmark to fix pipelines drowning in 50x more AI-written code — then cheerfully admits he can't explain CI, and delivers a surprisingly rigorous defense of betting on competence, integrity, and EV instead of technical depth.

Asked what he’s most excited about, Suds reveals he just incubated a continuous-integration company with Kleiner and Benchmark. The problem statement is sharp: engineers who wrote 1x code now ship 50x more to deliver the same features, and it’s breaking testing, merging, and shipping pipelines while CI costs explode. Then Jessica asks him to explain CI to the layperson and he laughs: “Look, I can’t really explain what CI is. But I can explain everything around it.” (Dave: “better than any explanation Sam gives anyway.”)

What could be a gotcha becomes the segment’s real content: a defense of judgment-over-expertise investing. Sam frames the problem — AI has washed out every traditional signal. YC taught everyone to pitch; now every deck has passed through 87 AIs and every company seems reasonable, while ‘founder quality’ is diluted because everyone and their mother wants to found a company. Suds’ framework: competence, integrity, and EV. Has this person been in the right talent networks to hire the best people? Have they sold to these customers, worked in the industry, earned unique insight into product and distribution? Can you assess they’re not lying to your face? Then the only remaining question is how big it is if it works — no technical depth required. He did a therapeutics deal knowing no biology. Dave’s blessing is real: “this sounds like great seed investing to me.” And Suds’ kicker is the era in one line: “It’s fine, Claude knows about it.”

Key points

  • The incubation: a CI company with Kleiner and Benchmark. Problem: 50x more code shipped for the same features is breaking testing/merge/ship pipelines and exploding CI costs.
  • Suds can't explain CI to a layperson and doesn't care: he can explain everything around it — the problem, the founder, the market.
  • Sam's diagnosis of why signals are dead: YC taught everyone to pitch, every deck passes through 87 AIs, every company now 'seems reasonable.'
  • Suds' framework: competence, integrity, EV — right talent networks, earned industry insight, not lying to your face, and a big enough prize if it works.
  • He applies it consistently: did a therapeutics company knowing no bio.
  • The kicker: "It's fine, Claude knows about it" — technical due diligence as a solved problem.

Where they landed

SudsInvest on the person and the problem, not the mechanism. Competence, integrity, EV; the model can hold the technical details.
SamSympathetic: team-and-theme investing is right, but AI has washed every signal, which makes the vibes harder to trust than ever.
DaveFully endorses: that's just what great seed investing has always been — and Suds explains it better than Sam does.
JessIf you're pitching it, you should be able to explain it — rule number one. Amused but unpersuaded by the vibes defense.

Quotes

“It's more like just competence, integrity, and EV. Has this person been in the right talent networks to hire the best people? Have they sold to their customers?”— Suds
“Every company seems reasonable, right? Because it's all passed through 87 AIs at this point before you see it.”— Sam
“Probably should know more about CI or be able to explain it. It's fine — Claude knows about it.”— Suds

Suggested tweets

Engineers used to write 1x code. Now they ship 50x more for the same feature and it's breaking every CI pipeline. There's a company in that.

Tweet this →

“Every company seems reasonable now, because it's all passed through 87 AIs before you see it.” — Sam Lessin on why deal signals are dead

Tweet this →

Incubated a CI company. Can't explain CI. “It's fine — Claude knows about it.” Seed investing in 2026, honestly stated

Tweet this →

All topics · Full episode #165 + transcript · Subscribe on YouTube · Spotify