Convergence
Three operators converge on a shared economic claim: AI agents change the unit of value from per-seat software to outcome-based labor substitution. The benchmark question is the Economic Turing Test — would you hire the agent if you didn't know it was a machine? — which forces outcomes pricing and rewards top performers (and underfunded teams) disproportionately.
Operators
- Benjamin Mann (Anthropic) — The Economic Turing Test — would you hire the agent if you didn't know it was a machine?. The benchmark: would you hire it if you didn't know?
- Bret Taylor (Sierra) — Agents push SaaS from per-seat to outcomes-based pricing; the incentive flip changes everything. Agents push SaaS from per-seat to outcomes-based pricing; the incentive flip changes everything.
- Boris Cherny — Underfund teams deliberately so AI substrate, not headcount, absorbs the work. Underfund teams so AI substrate, not headcount, absorbs the work — the operational expression of the rev-per-employee bet.
- Sherwin Wu — AI tools widen the spread between top and bottom performers — invest in top performers. AI tools widen the spread between top and bottom performers; invest in top performers.
Variation
- Mann provides the benchmark question.
- Taylor provides the pricing-model implication.
- Boris provides the org-design move (underfund deliberately).
- Sherwin provides the talent allocation rule.
- Convergence: the unit shifts from seats to outcomes; the org shape compresses; the top operators run away.
Implication
For application companies: model your ARR roadmap with outcome-priced agent revenue, not seat-priced software revenue. For team leaders: deliberately undersize teams so substrate absorbs work; over-invest in the top quartile of operators. For evaluation: ask the Economic Turing Test on every agent build — if the answer is no, the agent isn't shipping value yet.
Sources
- ins_economic-turing-test-for-ai — Benjamin Mann
- ins_outcomes-pricing-restructures-saas — Bret Taylor
- ins_underfund-deliberately — Boris Cherny
- ins_top-performers-benefit-disproportionately — Sherwin Wu