# 20VC: Mercor CEO on Why Application Layer Companies Have No Defensibility, The Model is the Product | Token Spend Will Exceed Headcount Spend in 5 Years | The True Cost of Hiring AI Researchers in the Valley Today with Brendan Foody

Page: https://stenobird.com/podcast/twenty-minute-vc/20vc-mercor-ceo-on-why-application-layer-companies-have-no-defensibility-the-model-is-the-product-token-spend-will-exceed-headcount-spend-in-5-years-the-true-cost-of-hiring-ai-researchers-in-the-valley-today-with-brendan
Text version: https://stenobird.com/podcast/twenty-minute-vc/20vc-mercor-ceo-on-why-application-layer-companies-have-no-defensibility-the-model-is-the-product-token-spend-will-exceed-headcount-spend-in-5-years-the-true-cost-of-hiring-ai-researchers-in-the-valley-today-with-brendan.md
Podcast: [The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch](https://stenobird.com/podcast/twenty-minute-vc)
Published: 2026-06-01T07:07:00+00:00
Episode link: https://thetwentyminutevc.libsyn.com/20vc-mercor-ceo-on-why-application-layer-companies-have-no-defensibility-the-model-is-the-product-token-spend-will-exceed-headcount-spend-in-5-years-the-true-cost-of-hiring-ai-researchers-in-the-valley-today-with-brendan-foody
Audio file: https://traffic.libsyn.com/secure/thetwentyminutevc/Brendan_Foody__Mercor_2.mp3?dest-id=240976
Processing state: not_requested
JSON: https://stenobird.com/v1/public/podcasts/twenty-minute-vc/episodes/20vc-mercor-ceo-on-why-application-layer-companies-have-no-defensibility-the-model-is-the-product-token-spend-will-exceed-headcount-spend-in-5-years-the-true-cost-of-hiring-ai-researchers-in-the-valley-today-with-brendan
Duration seconds: 4523

## Resource

Brendan Foody is the Founder and CEO @ Mercor, one of the leading data providers to the largest labs on the planet including OpenAI. In the last two years, Brendan has scaled the company to $1.5BN in ARR and a valuation of $10BN. AGENDA: True or False: Mercor lost Meta and OpenAI as a customer with the hack? Mercor has been poaching competitor talent, paying them millions? Mercor revenue is not real revenue and is only GMV? 12:56 Would Brendan sell Mercor for $30 billion? 14:23 Why everyone is wrong that AI will lead to labor displacement? 15:59 We will create many new jobs that do not exist with AI. 16:59 Why training agents will be a massive labor category that does not exist today 19:51 Will we see the data provider market unbundle and specialize into verticals? 22:24 Is the stated revenue really revenue or is it really GMV? 27:55 How a 1 million ARR company secured one of the best investors in the world with a helicopter ride 29:41 How Felicis secured the deal of the decade with a race track and a set of Ferraris 32:59 Which investment round felt like the highest price to grow into? 34:49 Why will value accrue to the infrastructure layer, not the application layer, in the next 12 months? 35:46 Why the model is the product and why application layer companies should be scared as a result 37:22 Why network effects will be the determinant of value creation 38:46 Why the forward-deployed motion, not the GTM motion, will determine true value creation. 41:59 Why token spend within organizations is going to continue to increase 43:54 Why agent evaluation to commoditize the model layer will be a massive business for enterprises? 51:13 Why we should have increased capital gains tax 01:01:31 How to compete with $20 million a year from Meta? 01:08:49 Will Mercor go public and…

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/twenty-minute-vc/episodes/20vc-mercor-ceo-on-why-application-layer-companies-have-no-defensibility-the-model-is-the-product-token-spend-will-exceed-headcount-spend-in-5-years-the-true-cost-of-hiring-ai-researchers-in-the-valley-today-with-brendan/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/twenty-minute-vc/20vc-mercor-ceo-on-why-application-layer-companies-have-no-defensibility-the-model-is-the-product-token-spend-will-exceed-headcount-spend-in-5-years-the-true-cost-of-hiring-ai-researchers-in-the-valley-today-with-brendan.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.