"We spend more on tokens than salaries"
Brandon Foody, co-founder and CEO of Mccor, discusses the challenges of building defensibility in the AI software layer, rapid demand growth, and the economics of AI companies. The interview touches on revenue quality, AI spending exceeding employee costs, and the high expense of recruiting top AI researchers.
Summary
This transcript is an introduction to an interview with Brandon Foody, co-founder and CEO of Mccor, described as one of the fastest-growing AI companies with a valuation exceeding $10 billion and over $1 billion in revenue. The host frames it as the most revealing interview Brandon has ever done.
A central theme is the difficulty of building defensibility in the software layer built on top of AI models. Brandon acknowledges that as the broader industry recognizes 'the model is the product,' creating durable competitive advantages at the application or service layer becomes increasingly challenging.
The conversation also highlights Mccor's explosive growth, with demand reportedly doubling overnight, outpacing the company's current infrastructure capacity. The interview promises to address whether Mccor's revenue is 'real' revenue and what the business model looks like going forward.
A striking economic data point is shared: Mccor is currently spending more on AI tokens for internal agents than on employee salaries, reflecting a fundamental shift in how AI-native companies allocate resources. Additionally, the high cost of hiring elite AI researchers is noted, with compensation often reaching tens of millions of dollars in stock per year.
Key Insights
- Brandon Foody argues that building defensibility in the software layer on top of AI models is going to be incredibly difficult, implying that competitive moats at the application level are eroding.
- Foody states that the broader industry has increasingly realized 'the model is the product,' signaling a shift in where value is perceived to reside in the AI stack.
- Foody claims that Mccor is currently spending more on tokens for internal AI agents than on employee headcount, illustrating a dramatic reallocation of operating costs in AI-native companies.
- The host notes that services are being automated in real time, suggesting that traditional service-based revenue models are under threat from AI-driven automation.
- Hiring a high-quality AI researcher often costs tens of millions of dollars in stock per year, reflecting the intense competition for top AI talent.
Topics
Full transcript available for MurmurCast members
Sign Up to Access