How far can AI really go in the complex world of food formulation?
To test its potential, we decided to run a simple experiment.
We asked ChatGPT to suggest new ice cream flavour combinations. The results ranged from the predictable (Earl Grey and lavender) to the intriguing (black sesame and coconut) to the questionable (black garlic chocolate chip).
This small test was just for fun, but it raises bigger questions about food innovation. While AI can certainly generate novel ideas, can it truly understand the nuances of taste, texture and consumer preference that drive successful product development?
AI can accelerate the product development process
Remember the old days of product development? The endless trial and error, mountains of market research, the weeks of waiting for consumer feedback?
AI is changing all that. It's like having a thousand researchers working around the clock, processing vast amounts of data in seconds. And we're not just talking about numbers – AI can even analyse opinions and emotions.
Olam Food Ingredients (OFI) recently let AI loose on a feast of online recipes, restaurant menus and food websites. It found dragon fruit and sesame are taking Europe by storm, while lychee and yuzu are winning hearts in the USA. Meanwhile, Indian and Indonesian consumers are falling in love with Western classics like butterscotch and salted caramel.
UK supermarket chain Waitrose is also getting in on the AI action. They've just launched 65 new products based on AI insights, including some pretty adventurous combinations. Hot honey pork belly slices, anyone? Or how about a mac 'n cheese quiche?
An evolution, not a revolution?
While AI's data processing capabilities are impressive, its role in food development requires careful consideration.
As Edward Norder from OFI points out, AI might be great at number-crunching, but you still need human expertise to turn those insights into delicious products. Global food strategist Cyrille Filott agrees – in a recent Just Food article, he said AI is more of a helping hand than a complete kitchen takeover.
Think about what Unilever's doing. They used AI to analyse millions of flavour combinations when developing their Knorr zero salt veggie stock cubes. They also used it to crack the formula for vegan mayonnaise, figuring out that modified corn starch could replace eggs to maintain that creamy texture.
For the most ambitious companies, AI could become a valuable tool in the development of novel foodstuffs such as proteins and enzymes, accelerating the sustainable food movement.
AI can’t replace human creativity – yet
AI can identify trends and solve specific formula issues, but food and drink is a creative industry. Sometimes, products that should make sense just don’t resonate with consumers.
Remember Coca-Cola's Y3000? It was developed using AI analysis from a massive survey of tastes and trends. Yet the reviews weren't exactly glowing (T3 journalist Rik Henderson declared, "if this is the future, I'm happy where I am").
And while it's fun watching YouTubers like RachhLovesLife test ChatGPT's recipe suggestions, we're not convinced feta ice cream is going to be the next big thing.
The truth is, successful product development requires a delicate balance. You need to push boundaries while providing comfort. You need to understand not just global trends, but regional preferences too.
Take Coca-Cola in Mexico, for instance. They use cane sugar instead of corn syrup – not because AI said so, but because they understand the cultural importance of the local sugar trade.
Developing new products that consumers WANT to buy
We've barely scratched the surface of what AI can do in the food industry. The potential is enormous, especially when it comes to understanding consumer behaviour and preferences.
But here's the catch: understanding what consumers want and creating products they'll actually buy are two different things. As the saying goes, "knowledge is knowing that a tomato is a fruit. Wisdom is knowing not to put it in a fruit salad."
Will the world ever be ready for black garlic chocolate chip ice cream? Maybe not. But AI could help us pioneer new foods and techniques that make product development faster and more successful.
The key? Using AI's insights alongside human expertise – particularly from people who understand regional nuances and how they could impact your recipes.
Hooley Brown helps brands localise product formulas, packaging and labelling.
Our global network of experts understand consumers in each region and can develop products that meet their preferences while complying with local legislation.
Get in touch to find out more.
This article was published in November 2024 and updated in February 2025. Facts were correct at the time of writing.
