AI and Food?

Have you ever wanted to try something new to eat? Or figure out what you're eating? This may have been a problem 20 years ago, where you might have consulted a culinary expert, but not anymore…

I'll be looking at two AI-based models, one is called Gastograph AI and the other is called the "ONE". Both of these models perform different tasks but essentially allow the person or firm to gain a better understanding of food. Analytical Flavour Systems, a company founded in 2010, released the app called Gastorgraph in 2016. We could assume that the 'gasto' part of the title stands for gastronomy which is the practice of cooking and eating good food. Then the 'graph' part is really what the app does. It maps out the different flavours one can experience from certain food and plots it on a graph. Yes, a person's taste is subjective, but the app works by using food tasters to share their input on a variety of food. This allows AI to use the information and form the app's body of data. And so, when a person writes a review of the food, the app will recognise what flavours match up with it. But where I believe the app really stands out, is showcased in its flavour comparison section. Manufacturers can analyse the flavour profile of their products, to see which flavour profile excels in the market. And if the products are similar, manufacturers can also compare their competitors' flavour profiles to do a collaboration with them – or to make more profit!


An example of flavour profile comparison. Credit: Google


McCormick and Company, an American spices firm partnered with IBM to produce an AI platform called the "ONE". The platform allows product developers to dive into the world of flavours. Not only has it shown novel food combinations, but also has expanded the knowledge and creativity of the developers.

McCormick’s use of artificial intelligence highlights our commitment to insight-driven innovation and the application of the most forward-looking technologies to continually enhance our products and bring new flavours to market.
— McCormick Chairman, President and CEO Lawrence Kurzius

The AI was trained with a huge magnitude of data. Data that has been collected over 40 years, which is about consumer taste, and what the general population find tasty. There are new spice blends which have already hit the shelves and aim to enhance family favourites such as, 'Bourbon Pork' and 'Tuscan Sausage'.

So, while both of these AI models perform different tasks, they have the same goal in the long run. And that is to revolutionise the food scene. Yes, it can be said that AI and food aren't talked about as much as other topics. However, this is changing. With AI driving forward and unlocking more knowledge about the food that we already know of, I say that it will not be surprising to see many AI-based food models that we can use, in a couple of years.

Thumbnail Credit.