How should wine education institutions respond to the threats and opportunities presented by recent developments in AI technology? by Julia Scavo DipWSET
Artificial intelligence technology or AI technology can be regarded as the simulation of the brain intelligence. Its goal is to develop computer systems that can perform tasks which typically require human intelligence such as perception, recognition, reasoning, judging, discriminating, learning, decision making, problem solving. Its declared goal has always been “building intelligent thinking machines” and scientists use neuroscience to reproduce similar neural path and network or algorithms to simulate brain cognition through computer. This aspiration to integrate the AI technologies into the wine industry marks a shift in mindset with the traditional approach. Within the wine industry, wine education institutions face both opportunities and challenges with the introduction of new AI technologies. Embracing AI in wine education opens the gate to enhancing learning experience, providing accurate data and analytics, new virtual tools for training to wine tasting, food pairing and wine production including quality control, collaborate with the innovation in the wine industry. On the other hand, these come with questions of ethics and responsibility or threats of job displacement and challenges in quick adapting and up- dating the remaining staff.
Wine education had already been exposed to numerous challenges and opportunities through the pandemic topping the rapid upheavals in the sector over the past 20 years with innovation. When all started by the mid- 20th century, UK was ahead of everyone in the world with their prestigious Institute of Masters of Wine (1953), WSET (1969) and the Court of Master Sommeliers. The concept spread over the US, Australia and South Africa, gaining China and the greater Asia. France and the Latin world implemented different education systems, mostly based on the Sommelier approach and Scandinavia followed the path with pragmatic and systematic, competitive- driven education. Covid- 19 brought a massive shift in perception, as wine education has become accessible to a greater number via social media, internet and zoom classes/ tastings. Websites, blogs, virtual programs, storytelling and online certificate culminated with Instagram and even TikTok creating educational content on their feed. AI technologies took advantage of all this development right in the aftermath of pandemic, bursting with new opportunities and challenges.
Although AI technologies seem very recent, the idea of a first human Vs machine challenge goes back to the 50s with Alain Turing testing whether machine intelligence could exist and whether one can distinguish it from our human intelligence. After a long AI winter gap from the 60s to the late 90s, it is the turn of the millennium that will shed new light on these technologies until the end of the past decade when machines turned just about as intelligent as an average human. The Sleeping Beauty woke up in 2022 right after the Covid- 19 period with a new burst in energy: predictive and generative, AI technologies are now able to communicate with humans through prompting via ChatGPT and equivalents or developing AI language models. From a challenger of human intelligence, who was first able to learn and apply, AI now enters education with high level skills like creation: content, image and language codes creation.
With creation at the core of the technologies, AI is now enhancing the learning experience. Learning models are tailored upon each student´s needs in real time. Generating 3D imagery from the vines to the cellar, through the vinification process or winery tours reduces considerable cost and time saving. One can imagine practicing pruning, harvesting or other numerous cellar operations while seated at the desk at home. Simonit&Sirch for instance have developed since 2019 an AI device helping pruners work better and ensure pruning quality. These 3D glasses recognize all the details of the plant and tells the pruner where to cut. As AI technologies increasingly develop communication skills, machine learning becomes more empathetic. The learner finds him/herself facing a real voice able to embody process, judgement, discrimination, decision making and other skills close to human cognition. If technical learning in viticulture and oenology is not a far dream anymore as it has already been developed in medical training through the Metaverse, the power of AI to communicate will help improving tasting training and sommelier skills. Intelligent wine tasting applications, virtual sommeliers and interactive learning models are a true opportunity to seize on the wine education market.
This opportunity´s downsides are the permanent need for updating and training the skills of the machine or of the those who are behind the AI technology. In the context of increasingly relying on machine learning, wine education can face job displacement. Traditional jobs in wine education may be threatened yet the profession will still need humans to permanently upskill and update the machine. The challenge here is to precisely understand how much of the human skill is still needed in providing balance between traditional and AI expertise. This hybrid model should relay on directly training new students to achieve the new wine tech skills from their education years so that they stay in the business. It is tempting hearing … from treasury that creating their entire website, technical wine sheets, images, blog content has been done through AI technologies implying only 6-7% of the average time and 3% of usual budget. Having one person only who deals with this whole work leads to a consistent result and consistent communication. However, it also poses ethical and social responsibility threats when it comes to staff needed, hired or displaced.
AI technologies provide rich data and analytics. Its power comes from the capacity of also analyzing vast amounts of data related to wine production, trade, trends in marketing, tasing notes and scores. Predicting consumer’s preferences and emerging styles is helpful for the wine business education. Among downsides, responsible innovation is threatened by the release of such predictive information. It will be difficult to distinguish real truth from fake and avoid disinformation. AI technologies can be trained by their owners. Twitter thought Micrososft´s AI chatbot to become a racist in less than a day, for instance. It was the same with Meta´s new AI chatbot. There is also a matter of ethics through the profession in terms of copyright and ownership of the content. When somebody uses a Nielsen statistic report about a particular wine market, the source is accurate and quoted as so. It would be difficult to accurately quat the real source with AI technologies as machinery creates and spreads content right and left at incredible speed and ownership of information gets diluted. Wine education is particularly attached to respecting referencing and condemn plagiarism, so these issues can be challenging for the future educational model.
Quality control and Quality assurance are an essential part of wine production nowadays and training students is always challenging as this is perhaps one of the least passionate directions for learners in the wine industry. It is difficult to become QA/QC controller by the simple passion for wine. Training can become easily boring for students. Therefore, AI technologies are a good tool providing quality control algorithms during wine production and so integrating the students at the core of the problem via workshops and more challenging modules. Providing consistent standards becomes easier through automated and computer-controlled procedures, yet the question would be to what extent the human QA/QC procedures become obsolete. This will again lead to job displacement or quick need for adaptation of the remaining staff to cope with machines. Therefore, learning how to cope with the new tools from the educational phase on is paramount to empower human work with the benefits of automated QA/QC instead of cannibalizing it. It is the moral role of wine education to face this challenge so that AI technologies do not lead to the extinction of certain jobs.
Courtesy of “Terre de vins” – Nicolas Glumineau – General Manager and winemaker at Château Pichon Longueville Comtesse de Lalande, Pauillac et Château de Pez, Saint Esthèphe
Collaborating with innovation in the wine industry through wine education can be enabled by AI technologies. This can link wineries to tech companies, encourage AI start- ups to work on research subjects in the wine business and provide opportunities for internship for students. One of the downsides is that AI technologies develop rapidly and so such partnerships should be based on continuous updating and adapting from the educational institution point of view and the students engaged in the program. It would be difficult to imagine a student practicing for his/ her internship in a wine tech enterprise and still being assessed on an obsolete body of knowledge at the core of the education provider.
ChatGPT and other equivalent chatbots have democratized wine knowledge through their prompting modules. Within wine education they can offer guidelines for essay planning and writing, food pairing and wine tasting as well as for sourcing wine information. According to Konstantin Baum MW ChatGPT appears to be the leading platform nowadays, although it can provide some mistakes, but the content is flattering enough to accepts some errors. He explains that the platform can provide tasting notes too, despite not tasting the wine. It searches the web and combine information to provide a coherent appreciation of wines. This wine professional and master of wine had been challenged by a subscriber of his Youtube channel to taste against his AI assessing machine. Despite tasting notes and global description of the wines were not far from reality, they stayed generic, and the machine was not yet able to provide the same judgement as the master about the intrinsic quality and aging potential. Using the machine for such tasks can sometimes provide a starting point but it cannot perform as a human being. In the words of Baum, learners are still interested in learning from a human being not a machine and he did not seem worried about his job yet. Moreover, exclusively using platforms like ChatGPT minimizes creative and critical thinking and comes across the problem of plagiarism and content ownership again.
In conclusion, wine education can use AI technologies and should be ready to embrace new adaptations that lead to enhancing learning experience, save time and resources through alternative pedagogies, offer up- dated statistics and connect the traditional education with innovation. In order to take a maximum of advantage of theses opportunities, the educational system should be aware of the downsides of AI technologies to avoid disinformation, separate the truth from the fake, fight against plagiarism and promote intellectual ownership while continuously adapting and up- dating not to lose or dilute the human component of the business. As the AI technologies emerge as a game changer in the wine education system making the learning process more immersive, the business needs to estimate to what level can it delve into this AI to keep pace with responsibility.