AI could predict and automate people’s food orders, says Deliveroo

British food delivery company Deliveroo has said that AI-powered meal delivery services could predict and automate people’s orders by 2040.

In its new report, Snack to the Future, the delivery giant said that operators would be able to tailor meals to people’s individual mental, physical, social and environmental needs and goals.

The global personalised nutrition market is already valued at £14.5 billion in 2023 and is predicted to grow to £66bn by 2041.

“In a category like ours where we provide a massive catalogue of meals, there’s definitely a role that we can play in providing a discoverability function that fulfils this need,” said Carlo Mocci, chief business officer UK&I at Deliveroo.

“Say someone has 450 calories available for dinner, had a heavy lunch and has no space for more carbohydrates, we might recommend a new dish from a restaurant that will make their nutritional goals for the day.”

AI-powered grocery delivery services are already using this approach to shopping by helping to automate customers’ orders based on their preferences, which also helps to reduce household food waste.

Deliveroo examined design studio Nonhuman Nonsense’s concept, AI-Bert’s Fresh Place, envisaging a supermarket run by AI that shapes customers’ shopping baskets based on their life experiences while taking into account ecological and societal factors.

Another concept that might appear in the future is ‘Breathprints’, the study claimed.

“Not only could people’s breathprint be unique and uncopied compared to fingerprints, breath health diagnostics could provide deeper levels of insight into our diet preferences and needs,” said Sarah DaVanzo, agency leader and F500 corporate executive.

According to the report, people’s breath is as unique as their fingerprint. Compounds in exhaled air produce a unique and stable molecular autograph or ‘breathprint’ that could be used to inform the early diagnosis of disease, for instance.

‘Breathprints’ could be used to ensure that what people order and eat has the optimal impact on their individual mood, energy, strength, weight and longevity.

“It could be common in the future because people’s devices could be enabled with BreathTech so that ordering could be all the more personalised to each person,” added DaVanzo.

Meal delivery services could also offer AI that synchronise with someone’s Personal AI – something Deliveroo describes as a digital version of someone’s mind – to provide personalised meal recommendations based on their physiological and psychological state at any given time.

Georgie Barrat, technology journalist and TV presenter, said that a personal AI will know how someone has slept the night before; if their heart rate has been high; if they feel stressed; where they are on their monthly hormone cycle — “all these things will build up an accurate picture of them.”

Deliveroo also expects the metaverse to shape the future of delivery services and restaurant brands.

“I think ASMR-style food content in the Metaverse could expand, with the therapeutic ‘brain tingles’ people experience being amplified by smart glasses. The look, sound, smell of a crunchy food could become all the more crunchier in the Metaverse,” said Tom Cheesewright, Applied Futurist.

“For example, augmented artificial smell technology is also getting really good. Meal delivery companies in 2040 could provide tools like this on their platforms.

Can’t decide between a burrito or poke bowl? Imagine, when ordering on food-delivery platforms, you can virtually smell and taste the food before you order, helping consumers to reduce time deciding between a wide-selection of cuisines and dishes.

When it comes to discovering (new) food the report suggested there’s a massive gap that could be closed by the introduction of advanced simulation technologies.

“Our choices are still driven by human activity, and algorithms don’t yet truly influence people’s behaviours,” said Cheesewright.

“But when you’ve got a virtual or augmented food and drink environment, like a restaurant, people will be much more attracted and immersed compared to a 2D experience. Imagine a digital menu with three-dimensional representations of steaming hot food. If it’s a cold day and I see a steaming bowl of pho, then I’m like, yeah, let’s order now.”

The same goes for making family food moments all the more inclusive and enjoyable.

“I could easily imagine children in the future that don’t like to eat their vegetables, putting on smart glasses and thinking they are eating candy. Technologies like this could trick people’s brains into thinking they are eating something more adventurous, luxurious or indulgent,” added Kaave Pour, co-founder of Space10, IKEA’s independent research and design lab.

This tech could also trick people into eating healthier food by using smart glasses to picture something more desirable.