Subvolts take AI is learning to smell, opening doors for disease detection and new consumer tech. Jul 8, 2026 - Developers
- Learn about novel AI applications in scent prediction and how graph neural networks can model complex data.
- Business
- Explore potential new markets for AI in disease detection, emotion sensing, and consumer devices through olfactory intelligence.
This interview explores a unique AI frontier: teaching machines to smell, with potential applications beyond fragrance.
How AI Learns to Smell with Alex Wiltschko - #771
Audio streams from Sam Charrington. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
Alex Wiltschko, founder and CEO of Osmo, discusses building olfactory intelligence for computers on The TWIML AI Podcast. The episode explores the science of smell, mapping molecular structures to odors, and creating AI models that predict scent. Wiltschko details Osmo's work building a large olfactory dataset and how this technology could extend to disease detection and consumer devices.
This episode explains how scientists are teaching computers to understand smells, much like humans do. They are creating AI that can predict what a molecule will smell like, which could help detect diseases or create new kinds of products.
This episode offers insight into a novel AI application: teaching computers to smell. It's for researchers and developers interested in multimodal AI and new data modalities. Osmo's work on foundation models for smell and proprietary datasets aligns with frontier labs' push to expand AI's sensory capabilities beyond vision and text, potentially leading to new diagnostic tools or consumer products.
- HostSam Charrington
- GuestAlex Wiltschkofounder and CEO, Osmo
- Guest: Alex Wiltschko, founder and CEO of Osmo
- Show: The TWIML AI Podcast
- Episode Title: How AI Learns to Smell with Alex Wiltschko - #771
- Length: 60 minutes
- Osmo's Dataset: Built the largest proprietary olfactory dataset from scratch.
In this episode, Alex Wiltschko, founder and CEO of Osmo, joins the show to discuss his goal of giving computers a sense of smell and what it takes to build olfactory intelligence.
We explore the science behind smell, from the hundreds of olfactory receptors in the human nose to the challenge of mapping the relationship between molecular structure and odor, ensuring safety regulations are met, and building foundation models for smell. Alex explains how graph neural networks and advanced embedding spaces allow AI to capture the multi-dimensional structure of scents, grouping them into perceptual neighborhoods, and creating a machine learning representation that predicts how molecules smell.
We also cover how Osmo built the largest proprietary olfactory dataset from scratch to train a fleet of predictive models, and how olfactory intelligence could eventually power applications far beyond fragrance, including disease detection, emotion sensing, and consumer devices.
🗒️ Full show notes: https://twimlai.com/go/771.
- olfactory intelligence
- Alex Wiltschko
- AI that can understand and process smells.
- graph neural networks
- Alex Wiltschko
- A type of AI that works with data structured like networks or relationships.
- embedding spaces
- Alex Wiltschko
- Ways to represent complex data, like scents, in a simplified format for AI.
- foundation models
- Alex Wiltschko
- Large, general AI models that can be adapted for many specific tasks.
53 min1 h 30Why Jev Is Changing How We Build With AI with Diogo Almeida - #779
WithDiogo Almeida
1 h 081 h 05From Voice Agents to AI Avatars with Alexander Smola - #777
WithAlexander Smola
Questions people ask
- What is Osmo's goal?
- Osmo aims to give computers a sense of smell and build olfactory intelligence.
- How does AI learn to smell?
- AI uses graph neural networks and embedding spaces to capture scent structures and predict how molecules smell, trained on large olfactory datasets.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to Sam Charrington.