Medvi: The Altman threshold artificial intelligence and the dissolution of the entrepreneurial playbook

In 2024, OpenAI CEO Sam Altman disclosed that he maintained a betting pool with fellow technology executives on when the first one-person billion-dollar company would emerge. “Without AI,” Altman said, “that would have been unimaginable. And now it will happen.” By the time he made that statement, Matthew Gallagher, working alone from a house in Los Angeles, was nine months away from proving him right. In March 2026, Gallagher faced a set of decisions that would determine whether Medvi, the telehealth company he had built in under eighteen months with artificial intelligence tools, two employees, and no external capital, could sustain the organizational model that had produced it. Medvi generated USD 401 million in revenue in 2025, its first year, with a 16.2% profit margin. The company was on track to reach USD 1.8 billion in projected annual sales in 2026. It had 250,000 patients, seven contract account managers, and one full-time employee other than Gallagher himself: his younger brother. Three decisions required resolution. First, whether to raise venture capital or continue funding expansion from retained earnings. Second, whether the zero-headcount organizational model that produced this growth could extend across four new product verticals simultaneously. Third, how to assess the hold-up risk embedded in an architecture that outsourced every clinical, logistical, and compliance function to two third-party platforms. The decisions were not independent. Each implied a different position on what the theory of the firm looks like when generative AI removes the constraints that theory was built to explain.