KM4BI

We Gave Away $2,000,000 to 5 AI Builders | MOONSHOTS Live

Five teams built AI businesses in 90 days, spanning education, company formation, labor markets, tax recovery, and 3D assets.

Published

The Build with Gemini XPRIZE set out to show that a small team can identify a meaningful problem and turn it into a revenue-generating company in 90 days. The target was not a generic AI demo, but a solution capable of affecting at least 100,000 people.

Building a company in 90 days

Peter Diamandis frames the experiment as an alternative to the fading linear contract of degree, job, and career. His proposed path is to choose a problem worth solving, then combine AI tools with entrepreneurial agency to build a business and a future around it.

Cathie Wood argues that young people facing fewer entry-level roles can use the same approach: start an AI-native company, potentially with no employees, while continuing to interview. Even if the startup does not endure, it demonstrates initiative and practical AI fluency.

Where the judges see opportunity

Wood highlights healthcare and drug discovery, where AI could reduce failure rates, development costs, and timelines. Logan Kilpatrick identifies data as the next bottleneck for model progress, creating opportunities for companies with specialized datasets and domain expertise. Mark Pincus focuses on rebuilding social media with agents and product mechanics that reward useful interaction instead of passive scrolling.

Five focused products

Dodo Prep turns study material into personalized tutoring. Launchbridge takes a founder from an idea to a company with an LLC, website, and payments. MyFixum serves Nigeria’s informal artisan economy. Tishan Sierra automates VAT-credit recovery for Mexican businesses. Polyfork creates AI-native 3D assets that users can resize, recolor, remix, and customize.

Polyfork wins first place and $500,000 in non-dilutive capital, while Tishan Sierra takes second with $200,000. The broader lesson is that AI becomes valuable when attached to a specific workflow, customer, and measurable economic outcome.

What the experiment demonstrates

Small teams can now test markets, automate major parts of product development, and reach customers with less capital. Durable advantage, however, is likely to come from the combination of technology, proprietary data, vertical expertise, and distribution.

Source