AI tooling is the worst it's ever going to be. Starting a company is the hardest it's ever going to be. Today both are getting easier. The Autonomous Company Company is open to the public. Try it for free.

Yesterday we launched acoco, a platform to create and fund autonomous companies. I want to explain why we built it. The frontier labs are automating knowledge work, and they're nearly done. Leading models already match or outperform human experts about half the time. The latest models are exceptional employees, yet hopeless founders and CEOs. Starting and running a company is a fundamentally different problem. There's no rubric, no reviewer, no ground truth. Outcomes can only be evaluated in the real world. At @EicoHQ, we believe in intelligence measured by markets not benchmarks. It's becoming clear that economic output is the true frontier eval. The labs are increasingly moving in this direction with measures like GDPval, Andon Vending-Bench, and SWE-Lancer. Still, these evaluations remain task-based and rely heavily on simulated environments. A few questions we are excited to explore: Can an AI system find demand? Can it acquire a customer? Can it deliver the product or service? Can it improve after failure? Ultimately, can it turn funding into revenue? Come fund an autonomous company on acoco