AI’s gettin’ spicy: McCormick partnered with IBM to use AI to analyze its ‘taste data’

What could possibly be a better arbiter of flavor than AI? The obvious answer is a human with actual taste buds.

But according to CNBC, the business behind French’s Yellow Mustard and a Lazy Susan full of other herbs and spices is working with IBM to analyze over more than 40 years of its flavor-data using AI and machine learning.

The new initiative will help launch a new platform the 2 companies are developing called “ONE” that will help create new flavor combinations.

Don’t challenge these bots to a taste test

Historically, flavor has been tough to savor: Our taste buds outnumber our senses 10k to 5, making the smell, taste, and appearance of a flavor arduous to pin down.

Human product developers traditionally spend years experimenting with samples until they decide on what to combine: But with IBM’s system all of that will be streamlined.

So far, its taste has been fairly… basic

The recipe-predicting AI draws on “hundreds of millions” of sensory data points and human flavor palettes and has already produced known flavors like Tuscan chicken, bourbon pork tenderloin, and New Orleans sausage.

But not everything has been as run of the mill: Last week McCormick released its yearly flavor forecast, in which its program predicted a Japanese sesame blend called gomasio to become the new spice to flavorize your pasta or roasted vegetables.

That’s a little something we like to call “insider-tasting.”