Artificial Intelligence is influencing the alcohol sector in a big way. But just how is one of our oldest industries being transformed by our newest technologies?
To comprehend the scale to which AI could be impacting the sector in the (very much immediate) years to come, we can investigate its potential role in every stage of the alcohol consumption process. From generating the perfect brewing recipes, to pouring the perfect pint, AI and machine learning technologies are beginning to have a revolutionary effect.
In several leading alcohol companies, AI’s involvement is already evident from the very beginning of the creative process. For example, ‘Mackmyre’, a Swedish single malt whiskey distillery, has incorporated machine learning algorithms to help perfect recipes. They feed these AI models data inputs such as existing combinations, sales statistics and customer preferences which, in turn, output millions of original ideas for potential recipes that are likely to be popular. A (human) master blender then tastes the final product and either affirms the recipe as a success or adjusts the combinations where necessary. As such, this is a terrific example of how AI and human expertise can work together to produce a quality creative product.
Brewing the Product
When it comes to production, we can call upon AB InBev — a Brazilian-Belgian brewing company responsible for mainstream names including Budweiser and Stella Artois — to offer examples of how AI can lend a helping hand. They use state-of-the-art AI networks to help predict the quality of their final product. Their systems analyze real-time data collected throughout the brewing process -including the amount of CO2 used and the length of time each stage takes to reach completion — and feeds this to the network. The resulting beer is then graded by professional beer tasters and these scores are also input into the AI model. By repeating this process, the algorithm can learn the optimum conditions required to produce the perfect product and, after further repetitions, the model can finetune these settings to produce an ever-improving brew. Again, we can see how human-AI interaction is crucial, as the process is reliant on human expertise informing systems on how the process can be bettered and perfected.
AB InBev has also found a use for AI in the development of strong customer relations — customers being the outlets and distributors that sell their products to the public. They feed a separate machine learning algorithm information on how much a given outlet purchases, sells and doesn’t sell over a set period. Learning from this input, the algorithm can then inform on the amount of stock that should be delivered to each outlet as well as the payment terms that would be appropriate given the current season or weather conditions. Thus, stores are generally better stocked, wastage is reduced and both businesses benefit from AI’s involvement.
Buying the Product
In this modern age, and with the help of advances in AI technology, the way we purchase our alcoholic products is also changing.
Take ‘IntelligentX’, for example. They offer a subscription service whereby customers receive an initial set of 10 beers and complete a feedback survey (via an app) outlining what they did, and did not, enjoy about the drinks they received. The AI algorithm then processes this feedback, compares it to similar responses from other customers and incorporates external factors (such as production constraints or seasonal considerations) to generate personalised guidelines outlining the kind of beer they are likely to enjoy. This information is then passed on to brewers who develop a new recipe and brew an entirely personalised beer that will be included in the customer’s next delivery. This process is repeated each month so that the deliveries contain increasingly enjoyable, and ever-changing, refreshments.
So, what’s next?
Evidently, AI’s involvement is rife within the alcohol industry, but it has certainly not reached its peak. For one thing, ‘digital tasting systems’ are becoming ever more sophisticated so that one day we will see the opinions of human experts being combined with the entirely objective feedback from an AI tester so that beers and spirits can be generated with ever-increasing quality.
Additionally, we may see more changes in the way we are consuming beer. As well as seeing AI impacting how subscription services provide beverages to private homes, we are likely to see an influence of AI on how beers are being served in our local pubs. ROBOBEER, for example, has developed a robot that is informed by a machine learning algorithm so that it is capable of pouring the perfect beer every time — the right amount of foam, with perfect conditions that maximise taste.
Clearly, while AI has had a major impact on how our alcoholic drinks are being made, distributed and bought, there is still more to come regarding AI’s influence in the alcohol industry.
Digital Tasting Systems: https://link.springer.com/chapter/10.1007/978-3-642-25167-2_48
Uniform Pouring — robobeer: https://www.mdpi.com/2306-5710/5/2/33/htm
Originally published at https://www.linkedin.com.