Impact of AI on the Brewing Industry
With Open Al’s introduction of ChatGPT in 2022, generative AI was propelled into the mainstream with free, easy-to-use and readily accessible tools. While some see the meteoric launch of ChatGPT as the birth of AI, traditional AI tools have already been incorporated into the brewing sphere for the automation of some brewing processes. Analytic AI is being used to examine and provide insights on large, complex sets of data to predict consumer preferences and inform brewing trends.
Perhaps the most controversial application of AI, involves the newer generative frameworks for content creation, marketing campaigns, and artwork. In this article, I would like to explore a few different applications of AI technology in the brewing industry.
Raspberry Pi & BrewPi
First released in February 2012 as a low-cost, general purpose single-board computer, Raspberry Pi was quickly adopted by the brewing community for automated fermentation. The first official software, known as BrewPi was released later the same year, to read temperatures, control heating/cooling plates, and to enhance fermentation during the brewing process. Using smart sensors, AI-based systems track fermentation patterns, adjusting timing and temperature to achieve ideal conditions. BrewPi is not an automated brewing process, but a tool used by a brewer to effectively regulate temperatures during the fermentation process.
AI & Sensory Analysis of Beer
In 2024, scientists from Belgian’s KU Leuven Center for Microbiology published a 5-year study that combined vast chemical and sensory data from over 250 unique brews. Their findings were then used to train AI models to predict specific flavours and customer appreciation of different beer. For each beer, the team measured over 200 chemical properties, with a sensory description of each one by a trained tasting panel. In addition to lab-based data, over 180,000 customer reviews of different beer were introduced to the AI machine. The research team trained 10 different learning machine models, choosing the best performing algorithm to predict customer appreciation of beer.
The practical application derived from the study is that researchers were able to identify specific, often unexpected compounds, that contributed to a beer’s overall flavour and the customer appreciation of it. This study revealed links between specific flavour compounds and consumer perception, providing a framework to develop flavours tailored to consumer preference, though with a significant amount of essential human input.
Generative AI & Content Creation
In 2023, Beck’s became the world’s first brewery to release an AI beer, from a recipe created by AI, with its marketing design and label art generated through AI. The beer, named Autonomous by its AI creator, was released as part of a marketing initiative to celebrate Beck’s 150th anniversary, though the company has yet to release another AI brew.
Other breweries, such as Heineken, have expressed a desire to grow with AI technology, with an emphasis on data mining for managing consumer insights, marketing strategies, and product innovation. Tony Costella, Global Consumer and Market Insights Director at Heineken, has stated that they are taking a cautious, yet progressive approach to AI. According to Costella, Heineken is part of a larger conversation with regulatory bodies about developing guidelines for ethical use of AI within the beverage industry, particularly with regard to content creation.
AI-based content creation is free, easy to use, and can be used as a source of guidance, design template, or to create a final product. For some craft breweries this may be an enticing option, especially for those with small budgets for designing label art and creating marketing content.
The major issue surrounding AI-based content creation is that generative AI software is “trained” almost exclusively using intellectual property of writers, artists and photographers, largely without consent. The question of whether this constitutes intellectual theft is at the core of a highly contentious debate, though I would argue that it undermines the work of real artists. In addition, it is becoming abundantly clear that generative AI is growing at an astonishing rate, while regulations for its ethical use are desperately needed, and lagging far behind.
In terms of the AI resources I have discussed in this article, there’s still a very real and important human element required in all fields. Beer is still brewed by a brewer, a learning machine will never be able to “taste” beer better than a human, and AI-generated content is generic when compared to that of real people, who have made deliberate, careful and informed decisions about their work.
(Leah is a Toronto based freelance writer as well as the Beer Boss and a server at C’est What)