Key Takeaways

  • Senior engineers leaving Big Tech forego $100,000 to $200,000 in annual cash compensation, turning each technical hire into an internal check written in foregone wages.
  • Venture capitalists typically make investment decisions after three or four partner meetings, while top technical hires run weeks of backchannel diligence before accepting an offer.
  • Talent panel data revealed Anthropic was pulling higher academic and technical pedigree than OpenAI years before Anthropic's valuation spiked.
  • Engineer retention serves as a leading indicator of internal health; Anthropic maintained lower developer churn than OpenAI during competitive model releases.

The Foregone Salary Thesis

Most investors evaluate early-stage startups by scrutinizing early revenue, waitlists, or founder charisma. Jerry Ye watches employment data panels instead. When an elite engineer walks away from a guaranteed package at Google or Meta to join an unproven startup, that engineer is making an equity bet that dwarfs the average angel check.

“Each engineer or each product person you're adding that could have made an extra 100k, 200k per year at a Max Seven could be seen as like a mini round,” (play from 4:22) Ye says. “And in aggregate, that's actually a very strong signal in terms of conviction.” (play from 4:33)

David Weisburd highlights the diligence asymmetry between investors and operators: “The difference there you could argue is that venture capitalists maybe have two, three, four meetings with the team and somebody that's quitting Google to go join a startup is probably doing a lot more diligence, is probably talking to a lot more friends that work at this company.” (play from 5:07)

A venture partner spends five hours with a founder across Zoom calls before issuing a term sheet. A principal engineer spends fifty hours evaluating the codebase, grilling peers on Slack, and verifying whether the architecture actually works. Their career equity is on the line. When ten of those operators pool their labor into the same entity within six months, the technical diligence has already been completed by the market's most qualified reviewers.

Why Anthropic Showed Up Early in the Data

Long before capital markets recognized Anthropic as OpenAI's primary rival, public talent panels showed a divergence in recruiting quality. Tracking the resumes entering both organizations surfaced clear patterns in academic background and prior employer selectivity.

“In terms of what you would see on a profile of a person in terms of resume, it's, you know, what schools they went to, that's a signal for selectivity, what other companies they worked at, where they went to, right? How many people went there, what churn look like,” (play from 8:56) Ye says.

Tracking educational selectivity and prior employer quality revealed that Anthropic was quietly pulling ahead in technical density. “So I think anyone that was looking at talent signal a couple years ago would have seen that on average Anthropic was hiring people with, you know, more in terms of like a pedigree types signals, like better schools, better experience, and on average that actually had them surpassing OpenAI at some point.” (play from 8:10)

The second metric was engineer retention. Even as model capabilities appeared neck-and-neck to the public, internal movement told a different story. “And if you look at it from a churn perspective, and it might be a reflection of the evaluation increases now, but if you look at churn, Anthropic is churning engineers at a lot lower rate than OpenAI,” (play from 7:36) Ye says.

Startups bleed engineers when technical leads realize the underlying tech is smoke and mirrors. When churn stays flat while headcount scales against top-tier incumbents, the engineers are seeing internal breakthroughs that external benchmarks have not yet captured.

Why It Matters

Talent tracking strips the marketing narrative out of deep tech investing. In sectors where product performance is opaque and sales pipelines take quarters to materialize, labor flow is the only real-time metric that cannot be spoofed by public relations. When institutional capital chases headline valuations, tracking where technical staff commit their human capital reveals who actually owns the frontier.