Nowadays, a tiny, nearly embarrassing incident occurs in kitchens all over the world. With your phone in hand, you stand in front of the open refrigerator and ask an app what you should eat. This would have seemed absurd ten years ago. It hardly registers at all today. The app computes, the refrigerator hums, and a model trained on millions of meals determines whether your leftover rice is a suitable dinner on a remote server.
It’s difficult to ignore how swiftly this change has taken hold. In the past, nutrition was the domain of dietitians, grandmothers, and the occasional glossy magazine spread. These days, it’s complicated by wearables, big language models, and chatbots that converse with you like tireless, patient dietitians. However, it’s still unclear if all of this is actually improving people’s health.
A portion of the story is revealed by the numbers. There are currently about 165,000 mobile health apps available in app stores, and the market has grown to be worth about $40 billion. According to surveys, nearly 83% of registered dietitians say they now rely on these tools in their practice, and over half of American mobile phone users have downloaded at least one health-related app. The final figure is the one that lingers. Something has obviously changed when professionals begin to rely more on the software than their printed handouts.
However, there are persistent reasons why people eat the way they do. When an algorithm recommends grilled fish over biryani, hunger, taste, mood, the cost of tomatoes, and the aroma wafting from a roadside stall on a leisurely evening do not vanish. According to the European Commission, approximately 800 million people worldwide suffer from hunger or malnutrition, while millions more, ironically, consume excessive amounts of the wrong foods. Despite its cleverness, technology is awkwardly positioned in the middle of this mess.
It’s surprisingly good at noticing. After a big lunch, a wearable on your wrist monitors heart rate variability. You’ve missed breakfast four times this week, according to a diet tracking app. With over 90% accuracy, AI models trained on photos and food labels can now recognize what’s on your plate. They can read a packet of biscuits more quickly than most people can squint at the ingredients list. Nutrition, which has long been viewed as a matter of conjecture, seems to be finally becoming quantifiable in a way that feels more individualized than clinical.
Initiatives like WiseFood, a Horizon Europe project operating covertly throughout Slovenia, Hungary, and Ireland, are testing whether all of this data can genuinely alter household behavior. Their tools use language models that respond in everyday speech to encourage users to follow healthier and less wasteful recipes. In 2015, it would have sounded futuristic, but now it seems almost commonplace.
Still, there are uncertainties. A deluge of unverified nutrition claims, miracle diets, and self-assured strangers with ring lights can be found on social media, which is frequently the same channel through which evidence-based guidance arrives. The amount of negative information can overpower the positive. Though it’s still unclear whether the average person eats significantly better thanks to an app or just feels more watched, investors seem to think the digital nutrition space will continue to grow.
As one observes this, it seems as though technology is merely sitting next to the messy human relationship with food, taking notes, and making subtle suggestions rather than taking its place. The question that no one can yet answer is whether those nudges result in long-lasting change or subtly disappear like step counts once did. As of right now, the app keeps making suggestions, the refrigerator keeps humming, and dinner still needs to be decided.⁖※
