Brandon Foody states he would not sell Mercor for $30 billion because he is motivated to build a legendary independent company.

Brandon Foody believes that over the last two years, the industry has increasingly realized that the AI model is the product.

Mercor is currently spending more on tokens for its internal AI agents than on employee headcount.

Brandon Foody predicts an enormous boom in AI security engineering tools due to new threats like swarms of coding agents.

Mercor's relationship with OpenAI remains strong and has not been terminated, contrary to rumors.

According to Mercor's APEX index, the capability of frontier AI models has increased from a score of 1% to 40% over the last 12 months.

Brandon Foody predicts that within 5 years, the average enterprise will spend more on compute than on employee headcount.

Brandon Foody predicts that the API layer for foundation models will become commoditized due to zero switching costs.

Brandon Foody predicts that the majority of AI inference in 5 years will use open-source, custom fine-tuned, or distilled models rather than frontier models.

Three former Mercor employees have founded companies that are now worth over $100 million each.

Mercor plans to go public within the next few years but is not rushing to do so in the next year.

In response to a security incident, Mercor engaged Mandiant and other security consulting firms and added security as its 7th company value.

Mercor's business relationship with Meta is currently paused.

Brandon Foody predicts that AI models will be able to clone entire SaaS applications like Slack end-to-end within the next 12 months.

Brandon Foody predicts that Mercor's daily payout to its expert network will triple to approximately $9 million within the next 12 months.

Companies that lack network effects will struggle significantly as their software moats are not defensible against AI-driven replication.

Hiring a high-quality AI researcher can often cost tens of millions of dollars in stock per year.

Brandon Foody believes that all knowledge work is converging on training AI agents, as it is structurally more efficient than redundant human work.

Four months after its Series A, Mercor received a term sheet from Felicis at a $2 billion valuation when it was at $20 million in ARR.

Mercor has over $500 million in cash and operates as a profitable business.