How AI Learns to Smell

TWIML AI Podcast · rss · 2026-07-09

Osmo founder Alex Wiltschko shared progress on giving computers a sense of smell in a recent podcast.

The team utilized graph neural networks and advanced embedding spaces to map molecular structures to odors, successfully building a foundational model for olfaction. They also created the largest proprietary olfactory dataset from scratch to train their predictive models. In the future, this technology could be applied in fields like disease detection, emotion perception, and consumer devices.

Original post →

More from Research

Research channel →