Mozilla
Day 38 after signup
Call for proposals: AI no bigger than it needs to be.
Submit a proposal for Human-Scale AI Incubator. Apply by Nov. 20.
Why it works
- Open with a concrete analogy to make an abstract technology critique easy to grasp.
- Explain the program’s purpose before detailing its funding and example projects, so readers understand the opportunity in context.
- Make the application deadline and primary action explicit, while placing secondary links and donation asks later in the email.
How it is built
- Illustrative opening
- Problem framing
- Incubator overview
- Project examples
Email copy
One size does not fit all. Hello, You wouldn’t blow out your birthday candles with a fire extinguisher, or butter your toast with a machete. But for the tech giants building AI today, the products come in one size: extra, extra large. Big AI models and bigger data centers are better for their bottom line, and we’re all paying for it. Almost thirty years ago, we fought against a one-size-fits-all browser. Today, we’re facing a sequel with one-size-fits-all AI, and it has us asking: What if AI could be built at the scale of the problems it’s trying to solve, instead of just as big as possible? We’re investing in this alternative trajectory for AI. Our next Incubator cohort focuses on projects that are small, tailored, and intentional. We call it Human-Scale AI: purpose-built tech, sized for the problem it’s meant to solve. Seven projects will each receive up to $500,000 over three years. We'll bring together an unlikely mix of builders and projects from many countries and communities, so ideas can travel across borders and contexts. To seed the Incubator cohort with some great ideas and leading innovators, we've preselected three anchor teams we see as foundational thinkers and practitioners in Human-Scale AI. They are: Tiles: Personal agents like Meta’s Muse and OpenClaw are becoming AI’s next mass-market product, but many of them concentrate sensitive data and control in a single company’s platform. Tiles is testing whether an assistant can be just as useful when you own your data. Family Intelligence, Inc.: Smart home assistants like Amazon’s Alexa+ and Google’s Gemini for Home handle some of a family’s most intimate information. The Family Trunk keeps their information at home, not on someone else’s servers. Its pilot with low-income families will show others how to design for people tech often overlooks. DAIR Institute: Small and mission-driven organizations depend on cloud providers like Amazon Web Services, Microsoft Azure and Google Cloud, whose costs, control and environmental footprint often clash with their values. DAIR has built tools that let organizations run their own compute, and our funding will help make them easy for anyone to use. The best starting points are the places the big systems miss, the languages they don’t speak, the needs they weren’t designed for. Communities without signal, services without support, families without access to the knowledge they need. Sound like something you’re building? Dive into the details and apply by November 20, 2026. Read the call for proposals → Not the right fit for you, but perfect for someone you know? Forward this email along. Lindsey Dodson Director, Global Grants Mozilla Foundation P.S. The principles behind Human-Scale AI show up across our work, from student buildathons in India to conversations at Mozilla Festival. Want to go deeper and see what’s influencing our thinking? Start with Dr. Timnit Gebru’s “Frugal AI” and the Green Web Foundation’s “State of the Fossil Free Internet.” Mozilla Foundation is building a better technology future—powered by people, open by design, fueled by imagination. Connect with us Thanks for reading! You’re receiving this email because you subscribed to Mozilla Foundation. If you no longer want to receive our emails, we’ll understand if you unsubscribe. Give once → Give monthly → 1875 Mission Street, Suite #103-564, San Francisco, CA 94103 USA Legal • Privacy