For years, wine has undergone a transformation from a snobby upscale product to one which the masses have adopted. In traditional wine-producing countries, people are romantic about wine. In France, Italy, Spain and other countries, wine holds a spell among ordinary folks, wine drinkers and connoisseurs. But with advancements in technology, Artificial intelligence (AI) is now playing a role in the wine business.
The romance of wine in the United States started in the 1970s, when Napa Valley proved that they could produce world class wine. California wine, since then, became more than just table wine and expanded beyond growers and vineyards.
Wine in the 21st Century
Wine appreciation as a function of culture, has taken a turn towards the 21st century. Wine clubs started because of a popular appreciation for wine along with desire for wine education. Wine subscription services aimed to meet the demand for wine by people who do not have the time to go out and meet regularly to discuss wines.
Now AI has the ability to decide what wine you should drink. A monthly wine delivery system uses AI tochoose the right wine, for the right occasion – fitting a person’s preferences.
The success of online wine clubs and wine subscription services is growing as more neophytes sign up for their monthly wine recommendations. These services, like those offered by Weekly Selection, use trained sommeliers to choose interesting wines for their clientele. After the success of these online wine experts, newcomers bring Artificial Intelligence (AI) into the picture. Winc and Bright Cellars use AI to assist clients in deciding among thousands of wines.
Bright Cellars Lets You Explore Wine Options
Promoting the company with a “passion for wine”, Bright Cellars aims to educate the public with the end goal of helping the them discover wine. They make it easy to order wines, and easier to understand wine. The start of the process is in getting to know the customer.
With the use of a questionnaire on their website, they aim to capture the customer’s tastes. The questionnaire explores your likes, including your favorite chocolate, or how you like your tea. This is where the AI comes in.
The AI matches the resulting flavor profile with the wines. The flavor profile also learns additional information through user reviews and recommendations from the customer. With more online engagement, the AI learns and fine-tunes the profile, thereby providing better recommendations in the future.
AI for wine is not far fetched. It does not have the nose nor the palate for wine, but it depends on the subscriber’s preferences. Not every wine drinker is able to discern the subtle nuances of flavor and scent. However, most wine drinkers and enthusiasts are able to explain their wine preferences. They can explain the fruity aroma of a zinfandel, or the density and thickness of a sangria.
AI Helps You with Your Preferences
The AI’s questionnaire is about taste and flavor preferences. The territoriality of wine and its character is due to the soil and growing conditions of the wine. These characteristics of the earth and climate give wine its layers of scent and flavor. In lyrical terms, it is the distillation of the soil and its fruits. Smelling wine before putting the glass to the lips is a necessary part of the enjoyment of wine. It teases the palate, and the sense of smell to open up for more flavors to come. The initial whiff is an invitation with its layers of scents.
Character is further reinforced when the wine first touches the tongue. The profile is layer upon layer of various other flavors. There is the slight citrus taste reminiscent of oranges or apples, and the underlying tannin coming in the second or third wave, the woodsy flavor adds to the complexity. The flavor intensifies and unfolds itself during the whole process of drinking wine.
This is the essence of today’s AI engine: to bridge personal interactions between AI and the user. It is the same “personalization” strategy of Alexa, Google AI and Siri, that help the AI become more useful to the user.
The initial results of the matching might not be different even if there were changes between answers. However, this is just the starting point.
AI Does Not Do Wine Tasting
The problem with the use of AI is that it is not inherently credible in describing wine without the use of a gas chromatography-mass spectrometry. This tool will tell the user the composition of the wine. This composition, when compared to other wines, can enable exactly where the wine was grown and what year. At least in theory, it should be able to. This is a cold science for such a vibrant libation.
The discord occurs when using AI on food and drinks. However, it is not how AI is used in giving recommendations. A person has preferences about chocolate, food, drinks, activities, and others. These variables are the baseline in suggesting wine. The initial suggestion may not vary from one user to another, no matter if the preferences changed. What does matter is the succeeding wine suggestions evolve with the feedback from prior wine selections.
These recommendations help the AI narrow down the search. Sommeliers have tasted the various wines and given their comments. The AI matches these comments with the subscriber’s preferences and their reaction to prior wine offerings. It is an intelligent system based on an iterative process. It learns with succeeding iterations.
Succeeding monthly deliveries would show that the AI has learned something about the individual.
The service also allows the user to choose which wine he would like to buy, foregoing any recommendation.
Richard Yau and Joe Laurendi, who both went to MIT, founded Bright Cellars. They had tried other services and found that the random wine deliveries did not catch their palate. The iterative nature of their AI reflects this attempt at fitting a person’s taste. The Bright Cellars service also has a unique wine selection as their list of wines is not easily available anywhere else. This allows them to have better control of their pricing.
Winc follows the same general concept of Bright Cellars. It uses a questionnaire to get an idea of a person’s tastes. Then it provides a recommendation for a 4-bottle delivery at $13 per bottle. The customer has the option to upgrade from a Select list, replacing the base selection with more expensive ones, at a slightly more expensive price. Some wines cost up to $40.
The 4-bottle package is free of shipping and costs a minimum of $52. It comes with a booklet titled “A to Z of Wines,” which is informative and a perfectly good read providing information about wine quality, taste, acidity, body, as well as grape varietals like “Merlot” or “Pinot Grigio”. The booklet also provides suggestions about wine and food pairings.
There is no membership fee to pay. You can skip a month, and you can also cancel your subscription any time by informing Winc of the cancellation at least 2 days in advance of the delivery date. First time orders can avail of a $20 discount.
The additional cost that Winc charges is not for the wine nor the delivery, but for the service and the experience. This is an experiential service aimed at highlighting the wine service.
Like almost all wine club delivery service, it is possible to buy a better known brand at the wine store, however, the company has taken pains to ensure that the customer understands the wine and pairs it accordingly.
Wine subscription services are going high tech and they expect to grow in number due mainly to the success of Bright Cellars and Winc. Specifically, their success at instructing new wine drinkers on the finer things in life can make long time loyal customers.