The world of odor is immense, research suggests that the human olfactory bulb may be able to distinguish more than a trillion odors. Alex Wiltschko, Richard Gerkin and the team at Google have achieved a remarkable feat by mapping the scent of molecules and digitizing the sense of smell. By doing so, the team unlocks the potential to discover new odors and molecules. The team validated their model against new molecules, linked their findings to the biological mechanisms of odor, and expanded their findings to find new approaches to global health challenges. The real world applications of this transformational research is beyond exciting. Article in Scientific American here.
I’m psyched to see what the future holds in the world of smell.
General explainability in AI is very difficult — I believe that at this stage in the development of AI it is most effective to tackle explainability in context of specific domains/applications. Pranav Kuber and the team @ Piction Health have a great implementation of domain specific explainability - blog post here.
