As artificial intelligence becomes more integrated into daily life, many people are seeking out experiences that emphasize human elements—like taste, smell, and touch—areas where AI is not yet fully capable. While AI excels at tasks like data analysis and administrative work, it faces challenges in interpreting sensory experiences, which are deeply personal and subjective. This has led to curiosity about whether AI can truly understand and replicate human sensory perceptions, such as the nuances of wine tasting. At Salesforce’s Dreamforce 2026 event, the company showcased an AI agent named tAIster, which had been trained on approximately 400,000 different types of wines. This training included details about each wine’s appearance, texture, aroma, and flavor. During the demonstration, participants were given three glasses of unknown wines and connected to tAIster via a smartphone by scanning a QR code. The AI agent guided participants through a sensory evaluation, asking them to describe the wine’s look, texture, smell, and taste in detail. One participant, who admitted they were not a wine expert, described the wines as thoroughly as they could. tAIster correctly identified the first two samples but misidentified the third, guessing it was Veuve Cliquot champagne instead of Chardonnay. The error occurred because the participant mentioned seeing small bubbles in the glass, a characteristic more commonly associated with sparkling wines like champagne. Despite this minor misstep, the participant noted that the demonstration was still a proof of concept and not yet a threat to professional sommeliers. The experiment highlights both the potential and the current limitations of AI in interpreting complex human sensory experiences. While AI can process and analyze vast amounts of data, it still struggles with the subjective and nuanced nature of sensory perception. For now, it seems that human expertise in areas like wine tasting remains unmatched, though AI may one day offer valuable support or insights in such fields.