MIT's AI Project Invents New Flavour Combinations For Your Favourite Foods
Everybody loves pizza. Some people like to indulge in a nice margherita, while others like BBQ chicken. However, if you eat pizza a lot then you might get a little bored by the selection Domino’s has, so why not try a tasty shrimp, jelly and sausage pizza? Does that not take your fancy? How about a sweet potato, bean and brie pizza? These were some of the rather wacky combinations of pizza toppings that MIT’s new AI suggested.
This new AI is part of the “How to Generate (Almost) Anything” project which was started “to test the limits of human-AI collaboration to generate (almost) anything the human mind can (and can’t) imagine.” The pizza topping combinations were given by the AI and it even said to add non-existing ingredients like ‘snipped caramel cheese’ and ‘wale walnut ranch dressing’. My taste buds are already tingling!
The project’s YouTube channel showcases the other ventures in the scientists’ incredible journey. These include the human-AI collaborated theatre play, where an AI can write its own scripts, and human-AI collaborated perfume where an AI listed ingredients to make different perfumes. This goes to show that a lot can be done when humans and AIs work together so we can have the best of both worlds.
Unsurprisingly, AIs that are capable of generating flavours and food products are becoming something companies are looking to spice up their food game. McCormick & Company, one of the US’ largest food companies, announced it was collaborating with IBM to develop an AI just like this. Maybe companies like Calvin Klein would want to enlist the help of tech companies to develop a perfume creating AI like MIT’s so that people worldwide can smell the wonderful ‘Ventival’ or ‘Hivinga’ (perfumes generated by MIT).
Soon, human-AI collaborated chocolate truffles, graffiti, fashion and pretty much anything else could enter your life. In the future, when you walk past McDonald’s, the burgers may be designed by AIs and refined by humans. Maybe AIs could help design other AIs that design AIs and so on, but that’s a whole different story.