Building in Public as a Solo Founder in the Age of AI

Matt Paige argues that typing was never natural, humans speak much faster, and AI has made voice accurate enough to drive a shift from typing to talking with computers. He interviews Thomas Schlossmacher, who is building Resonant, a Mac voice dictation tool that runs entirely on-device so nothing goes to the cloud, motivated by privacy concerns and avoiding user data being used for training. They discuss why speaking enables more precise, contextual prompts; the awkwardness of dictating in public; and how Resonant differs from cloud-based tools like Whisper Flow. Schlossmacher explains challenges of running speech-to-text and LLM post-processing locally, the tradeoffs versus servers, and how future hardware may improve local performance. They cover product differentiation through taste and brand, his design workflow, his path to building with agents, his stack, and a vision for proactive, context-aware agent capabilities and potential open-sourcing. Resonant is available free in beta at https://www.onresonant.com/; Schlossmacher shares updates as sourceTMS. --- Key Moments: 01:47 Switching to Dictation 05:00 Why Build Resonant 08:25 Local LLM Reality Check 11:23 Standing Out in AI 14:52 Designing the Brand 19:16 Building Taste Systems 22:25 From Non Engineer to Builder 23:43 Early Web Hustles 25:07 Entrepreneur First and AI Shift 27:49 Learning Backend with Agents 29:05 Tooling and Tech Stack 31:05 Resonant Product Vision 35:14 Proactive Voice Workflows 36:42 Beta Launch and Monetization 41:15 Where to Try Resonant --- Key Links: Resonant: https://www.onresonant.com/ Connect with Thomas on LinkedIn: https://www.linkedin.com/in/thomas-schlossmacher/