Sommelier, hang in there. AI was taught to accurately assess the origin of wine

December 6, 00:23 Share:

AI is becoming a real competitor for sommeliers (Photo: wds_wine / Instagram)

Using machine learning, researchers have taught computers to distinguish truly high-quality wine from counterfeits of scammers. All the secrets are revealed by the chemical composition of the drink.

Scientists selected 80 wines from seven different estates in the Bordeaux region of France, produced over the past 12 years, and analyzed them using gas chromatography.. This method is commonly used in laboratories to separate and identify the compounds that form a mixture.

Algorithms analyze all the chemicals found in wine and generate a description for each sample, writes The Guardian. AI can not only determine the growing region of the grapes from which the wine is made, but also the specific estates where the drink was made, with 99% accuracy. At the same time, it is not easy for him to determine the age of the drink: the model showed an accuracy of less than 50%.

“There are a lot of wine scams out there: people making something up in their garages, printing up labels and selling it for thousands of dollars.”. We show for the first time that our chemical methods are sensitive enough to tell the difference,” said Professor Alexandre Pouget from the University of Geneva in Switzerland, who worked on the study.

While anti-fraud is the primary intended application of this technology, Pouget is also convinced that it can be used to monitor quality throughout the wine production process.

Read also: “Much smarter than us”. Musk said that there will come a time when artificial intelligence will take away jobs from people “Will treat you like animals”. A scientist from OpenAI warned about the threat of artificial intelligence to humanity Technologies against disasters. Artificial intelligence helps predict the appearance of rogue waves