Osmo, an AI startup spun out of Google Brain in 2023, says its smell model has reached "human-level olfactory intelligence." Founder Alex Wiltschko defines this as performing about as well as the average person at detecting and discriminating smells, and about as well as a junior perfume designer at creating them. The company says the model is already used in the fragrance industry, and that last month it outperformed in-house perfumers at one of the world's largest consumer goods companies. Osmo's representative did not respond in time for publication.
Osmo says it has built a dataset of more than 10 million human labels and 800 million "scent tokens," which describe the molecules and concentrations in specific odors. A team of around 70 trained data labelers in the US and overseas writes descriptions using a standardized fragrance vocabulary. Wiltschko says the model could eventually support threat detection at airports or disease diagnosis. The report says the company has worked with Procter & Gamble on retail fragrances, with the Gates Foundation on insect-repellent molecules, and with StockX on detecting counterfeit sneakers. The Gates Foundation's $5 million investment and $3.5 million grant are reported.
Critics are not convinced. Smell scientist Avery Gilbert argues on his Substack that some of Osmo's marketing claims are undocumented and that its patent filings still rely on human sensory panels and perfumers rather than AI. Michael Schmuker, a former professor of neural computation who has consulted for Osmo, calls the technology an achievement. He says Osmo may still miss context, noting that the same molecule can evoke different associations depending on circumstances.