The Pentagon's AI Equity Play & Scaling Startups with AI Agents
This Week in Startupsgo watch the original →
the gist
Jason Calacanis and Lon Harris discuss the risks of the US government taking equity in AI startups like Fluidstack, while guest Tibo Louis-Lucas demonstrates how to manage multiple companies using AI agent stacks.
The Risks of Government Equity in AI
Jason Calacanis critiques the reported $5 billion Pentagon loan to Fluidstack, arguing that the government's move toward taking equity in private AI companies represents a dangerous shift toward nationalization. He contends that the government should act as a customer through competitive procurement rather than a venture capitalist. The core tension lies in the inevitable conflicts of interest: if the government owns equity in a specific AI firm, it creates a bias that undermines fair competition and market-driven innovation. Calacanis suggests that if the state wants to participate in the upside of national champions, it should utilize warrants—similar to the missed opportunity with Tesla—rather than direct equity stakes that distort the free market.
Scaling Operations with AI Agents
Guest Tibo Louis-Lucas, founder of Squad, shares his framework for running seven companies with a team of fewer than 10 people. His approach centers on "stacking" AI agents to automate repetitive tasks, such as affiliate marketing updates, lead qualification, and customer support. Rather than hiring full-time staff for every function, Louis-Lucas advocates for using specialized $20/month AI subscriptions to handle high-volume, low-complexity workflows. He argues that for solo founders or small teams, the goal is to build an "autonomous marketing team" where AI agents handle the execution, allowing the human to focus on high-level strategy and product development.
The Evolution of the Startup Stack
Louis-Lucas emphasizes that the modern startup stack is no longer just about SaaS tools, but about the agentic loops that connect them. By integrating AI agents into existing communication channels (like Slack or email), founders can create a self-sustaining loop where the AI identifies opportunities, drafts responses, and executes tasks without constant human oversight. This shift allows for unprecedented leverage, where a single operator can maintain the output of a much larger organization by acting as an architect of automated systems rather than a manual worker.