AI Helps Battle Rancid Scourge

AI Helps Battle Rancid Scourge

AI Helps Battle Rancid ScourgeAs AI embeds itself further into our culture—and becomes increasingly better at mimicking people—humans in turn are finding new ways of making it work for us. Now, a team from Clemson University has harnessed AI to avert one of humanity’s peskiest culinary disasters—rancidity.

Foods go rancid when the oils or fats in them are exposed to oxygen, breaking down into unpleasant-tasting component molecules. (If you’ve ever chomped on a too-long-opened bag of chips or nuts, you know that particular metallic grossness.) Companies routinely add tiny amounts of antioxidants to foods that are prone to oxidization, in order to combat the process and preserve freshness. Sometimes it can be too much of a good thing: Certain antioxidants, in combination, cancel out each other’s effects. But figuring out when this happens and in what combos is a herculean task—one that authors Carlos D. Garcia and Lucas de Brito Ayres have figured out is perfect for AI.

“… We selected a type of AI capable of working with textual representations, which are written codes describing the chemical structure of each antioxidant. […]

Once the machine could recognize general chemical patterns, like how certain molecules react with each other, we fine-tuned it by teaching it some more advanced chemistry. For this step, our team used a database of almost 1,100 mixtures previously described in the research literature.

At this point, the AI could predict the effect of combining any set of two or three antioxidants in under a second. Its prediction aligned with the effect described in the literature 90% of the time.”

The team trained the AI further on 200 examples from their own lab, which netted them predictions that were more accurate to their real-world experimental results. The team likens an AI like theirs to “a very clever assistant,” which can run boring or monotonous tasks for its human handlers, and spit out the results they need. In helping solve the problem of rancid food, this AI/human pairing might just be our model for collaboration in the future!