Episode
08: The Language of Scent: How Alex Wiltschko Is Digitizing Smell
- Published
- Sep 8, 2026
- Duration seconds
- 3092
- Processing state
not_requested
Actions
POST https://stenobird.com/v1/public/podcasts/essential-ingredients-podcast-1367183/episodes/08-the-language-of-scent-how-alex-wiltschko-is-digitizing-smell/transcription-requests
Idempotently request low-priority transcript generation for this episode.GET https://stenobird.com/podcast/essential-ingredients-podcast-1367183/08-the-language-of-scent-how-alex-wiltschko-is-digitizing-smell.md
Read the agent-friendly Markdown representation of this episode resource.
Summary
"We want to build a platform and a fundamental capability that makes the world smell better… and we know from the fundamental understanding of how scent works that getting better at scent that delights you actually helps us get better at applications in scent that can protect and can heal." —Alex Wiltschko Digitizing smell sounds like science fiction until you meet someone who's spent two decades chasing it. In this episode, Justine Reichman sits down with Alex Wiltschko, founder and CEO of Osmo, a company on a mission to give machines a sense of smell — and, eventually, to help detect disease before symptoms show up. Alex's path started long before Osmo. A PhD in olfactory neuroscience, five years leading Google's digital olfaction team, and a personal loss — his father's death from brain cancer — all converged into a single question: what if scent could tell us things about our health that we can't yet hear? This conversation covers how Osmo's AI model, OI One, is learning to create and detect scent, why the company deliberately started in fragrance instead of diagnostics, and what digital olfaction could mean for the next decade of human health. This episode is for anyone curious about AI's next frontier, the science of scent, or what it looks like to turn personal grief into a decades-long body of work. Key takeaways: Scent is built from a surprisingly small alphabet — roughly 20,000 to 30,000 molecules make up every smell on Earth. Osmo intentionally avoided starting with disease detection because the science wasn't ready; fragrance became the training ground that builds toward harder problems. Osmo's AI tool, Osmo Studio, now generates starting scent compositions that professional perfumers treat like a collaborative apprentice. Synthetic fragrance ingredients ca…