Wineink: Artificial intelligence and wine

It has been quite the week for artificial intelligence, or AI as we call it.

Globally, the story was how a Chinese AI company called DeepSeek sent shockwaves through the tech financial markets when it was reported that it created its platforms for considerably less money than the industry leaders, like OpenAI. The chip producer Nvidia lost nearly $600 billion with a “B” in market capital in a single day after word came down about the development.

For perspective, the global wine market in 2023 was valued at $330 billion. At least, that is according to the “AI Overview” of articles that pops up when I Google “size of global wine market.” I have no reason to doubt that, and, in this day and age, if a result comes under an “AI Overview,” then we generally regard it as gospel.

---

---

Locally, the AI story was that the technology was being used at the X Games to see how it might improve the judging in competition. An AI application was developed in collaboration with Google Cloud for the SuperPipe event that analyzed camera angles to determine amplitude and the successful completion of a trick. It seems that AI can better calibrate the number of spins, twists, and turns a snowboarder makes than the naked eye of the judges. One of Google’s founders, billionaire Sergey Brin, even came to Buttermilk to present the medals to the Pacifico Men’s Snowboard Big Air competition winners.

What does this have to do with wine?

---

---

Well, the point of AI is all about improving business efficiency and the pursuit of perfection. And as these are both important goals in the world of wine, it should come as no surprise that AI is fast becoming a player in the production, distribution, and marketing of the wines we drink.

We like to think of wine as a quaint, handmade product made by artisans in a “natural” way. And for some wines, it is. But like any and every industry, technology and AI are leaving their mark on wine.

Farm-to-table or grape-to-glass: Those are three-word descriptions that define the wine industry. The product is made by nature in the vineyards, harvested by man, enhanced, and bottled by winemakers and ultimately sold by marketers. Each facet has been touched by AI technology.

In the vineyards, data is critical for getting the most out of a vintage. Numbers — including how many vines per acre, how many clusters per vine, how much rain or irrigation is needed, current sugar levels, and myriad other data points — were once kept by hand, then in computer spreadsheets. Today, AI learning tools are relied on to not only keep track of the of all details in vineyards, but also to create models outlining what the wines will taste like based on that information.

In 2024, Concha y Toro, the Chilean-based Latin American producer, launched an initiative with the French Institute for Research in Science and Digital Technologies (Inria) called Harvest 5.0, which aims to “develop a digital system integrated with predictive models based on explainable artificial intelligence to predict harvest volume, vine phenology, grape maturity monitoring, and the scheduling of grape reception and winemaking operations.” That pretty much covers it.

Third-generation farmer and wine producer Tom Gamble farms nearly 200 acres of prime Napa Valley vineyards and, in his own folksy but sophisticated way, became an early adaptor to AI back in 2022. The goal of Gamble wines is “to farm flavorful grapes at lower sugar levels that result in lower alcohol wines.” Technology helps fine tune that goal. Using electric, self-driving smart tractors made by Monarch that incorporate AI, along with drones and data collection, the winemakers can be more predictive about harvest dates and resource needs to achieve their goals.

Once harvested, the impact of AI technology in terms of selecting grapes and making wines is perhaps even greater in the wineries than it is in the vineyards. While optical sorters have been in use for a couple of decades now, the computer programs that operate them have become more advanced with the use of AI. Grapes that were once hand-sorted are now placed on conveyer belts and go under the eye of optical filtering systems. These select the most desirable grapes, based on color, size, and the condition of the grapes as determined by the winemaker. When fermentation begins, much of the guess work has already been taken out of the process, but, even here, AI-driven scripts can make minute changes in conditions as the wines evolve.

And of course, marketing wines is a science unto itself that is mostly driven by data on past purchases and expected market conditions. Like any consumer product, the sale of wines can benefit from information that is easily and clearly processed. Be it the stock market, sports judging, or the wine industry, the effects of AI are just now beginning to be felt.

For wine, it is paramount that the human touch, the nuanced skills of winemakers, and the terroir where the grapes are originally grown remain the most important components of what goes into a bottle of wine.

Wine is life. Let us not forget that.