AI - like having a crystal ball for flavour trends

Ahead of Flavorcon 2024 in October, we take a look at one of the most exciting innovations in flavour creation: Artificial Intelligence (AI). Flavorcon brings together flavourists and food product developers to explore the latest trends and advancements in flavour enhancement.

This year’s theme, ‘Unleashing the Future of Flavours,’​ delves into how AI is reshaping the flavour space. It explores the role of the flavourist in the context of AI’s growing influence, the swicy​ (sweet and spicy) trend, and market insights for flavour innovation.

AI flavour creation today​

“AI use in flavour creation for confectionery is becoming more common by the day,” says Somsubhra GanChoudhuri, CEO and Co-founder of Ai Palette, an AI-based consumer insights platform. Although still in the early stages, more brands are waking up to its potential.

AI’s capability to analyze massive amounts of data—such as social media trends and consumer reviews—allows it to predict flavour profiles that will resonate with specific audiences. “It’s not a niche thing anymore. AI is being integrated into the product development process in ways we couldn’t have imagined a few years ago,” says GanChoudhuri. “While it’s not yet ubiquitous, we’re right on the cusp of it becoming the new standard.”

However, AI’s role in flavour creation has its limitations. “AI on its own doesn’t bring any value,” says Jonathan Raemdonck, sales and marketing director at Foodpairing AI. He emphasizes that while AI can generate trends, combining scientific data with consumer preferences is key to successful product development. AI alone cannot magically mash together seven trends to create a successful product—it’s about a blend of science, creativity, and AI insights.

Pushing AI potential in flavour creation​

One of the standout innovations AI offers is the ability to sift through vast data sets—ranging from social media to taste reviews—and identify emerging flavour trends much faster than traditional methods. “Digital twins, knowledge graphs, and large language models (LLMs) enable rapid concept generation with new flavours and validation,” says Raemdonck. This technology can be used for new product development, line extensions, and even optimizing portfolios for market exploration.

A breakthrough feature of AI is its capacity to simulate ingredient interactions, predicting the flavours they create. This accelerates R&D processes, allowing real-time concept testing and helping brands gauge the success of new flavours before they hit the market. “It speeds up the R&D process and allows for real-time concept testing,” adds GanChoudhuri.

Combining AI with sensory science also allows for unprecedented innovation. By digitally modeling human sensory experiences, AI refines flavours based on generated feedback. “It’s a game-changer,” says GanChoudhuri. “It helps bring products to market more quickly and effectively.”

Challenges in AI-based flavours​

Despite its promise, AI faces challenges in the flavour industry. "When people think of AI, they often consider general platforms like ChatGPT or Gemini,” says GanChoudhuri. However, these models are not trained for the fast-moving consumer goods (FMCG) sector, and the challenge lies in ensuring that AI outputs align with sector-specific needs.

Flavour chemistry is complex, and AI needs access to high-quality data for accurate predictions. “Access to structured data is critical for AI to make accurate predictions, and the industry is still catching up in terms of digital transformation,” says Raemdonck.

Another hurdle is algorithmic bias and data hallucination—where AI generates insights that aren't grounded in real data. This can affect prediction accuracy. Furthermore, the complexity of flavour itself presents an issue. Taste perception is subjective, influenced by culture, biology, and individual experience, making it difficult for AI to generalize across global markets.

While AI can model sensory experiences, capturing the full range of human taste and smell remains challenging. “This limitation affects flavour precision,” says GanChoudhuri, which impacts the accuracy of real-time concept testing.

Additionally, implementing AI technology requires significant investment in both resources and expertise. For smaller brands, these costs can be prohibitive, limiting their ability to fully leverage AI in their product development processes.

AI’s impact on flavourists​

AI isn’t replacing human expertise—it’s enhancing it. “AI empowers flavourists to elevate their craft,” says GanChoudhuri. By leveraging AI-generated insights, flavourists can fine-tune product creation to a level of precision previously unavailable.

“AI needs to complement, not replace, human expertise,” says Raemdonck. By integrating AI insights into intuitive workflows, companies like Foodpairing AI ensure that flavourists retain creative control, while using AI to inform and enhance their decision-making.

In short, AI is reshaping the flavour industry, but it’s doing so as a tool in the hands of experts. Through a combination of data-driven insights, rapid R&D, and enhanced creativity, the future of flavour development looks brighter—and more exciting—than ever.