Solo Founders of AI Cooking Apps - Bridging Generations & Recipes

For nearly three decades, Matt H. has been steering technology in service of craft: from his college days wrangling early versions of Maya, to motion design at Buck and Psyop, to mixed reality installations at Viacom, to eight years at Meta scaling Spark AR and shaping Quest 3 and Orion. That range and experience show up in how he decides what AI should and shouldn't do. In today's newsletter, I sit down with Matt to hear about his work as a solo founder launching Heirloom, an AI-powered cooking app that bridges generations, from preserving your grandmother's recipes to helping your kids replicate what they saw in a 45-second cooking TikTok. Matt and I worked together at Meta back in the day, which made this conversation extra fun: a chance to catch up and nerd out about designing with AI. You'll learn how Matt: 🔍 Decides when to “hire” AI using Jobs to be Done. Matt first identifies the underlying Job to be Done for which a person will “hire” a solution. He then asks if AI is the right hire for the context of use, considering capabilities and stakes. ⚖️ Determines trust level first. Matt thinks about worst-case scenarios, and designs to that specific reliability bar. If the technology can't currently meet it, he finds a problem space where the stakes allow for more flexibility. 🎩 Embraces “hat consolidation” without compromising quality. AI collapses the distance between disciplines, allowing you to wear multiple hats. That isn’t an excuse to lower your quality bar: just because you can now “write” code isn’t an excuse for making slop. This episode continues one of my favorite traditions from writing my Rosenfeld Media book, Designing Automated Futures: conversations with practitioners navigating AI in their work. Know someone who should be featured? Drop me a line. 👇 Link to episode and Heirloom in the comments.