Science says that scent is the strongest sense to evoke memories, trigger mood or emotions. All along since ancient times, the craft of making a perfume hence has been deeply human. After all, it takes a human to understand another human being. Well, till now.
IBM is now deploying artificial intelligence to develop scents. A machine learning developed by IBM studies the existing fragrances, it’s chemical constituents and layers it over with existing data sets of geography and age groups.
The algorithm is named Philyra. It can now create new perfumes for targeted markets. Symrise, a global major in fragrances has tied up with IBM for this project.
Symrise has provided the valuable data for the project which included not only the 1.7 million fragrance formulas, but also the sales performance report of various scents, demographics and customer purchase insights. The algorithm then used this data, to create new scents.
Already two scents, created by AI have been sold to a leading Brazil based large retail store. Both fragrances received a huge response from the elicited focus group and were top sellers, in spite of competing with other better knows ones.
While the new fragrances created an be definitely enticing, but losing the human touch in an intimate process like applying scent, many would not want.
The good thing is that it can launch many new Perfumers, not self-styled but AI styled. The coveted and elite art may no longer be so.
Just like an apprentice would learn from a master what combinations of ingredients would work well, like what would be a good substitute for orange oil and when, or when to add rose oil instead of lemon, machine learning will create a fragrance based off which formula worked best.
Philyra’s creators believe this is one way of telling that, computers can definitely be creative!