it is adapterA weekly newsletter sent every Friday from the Verge senior reviewer victoria song It analyzes and discusses the latest phones, smartwatches, apps, and other devices that claim they are going to change your life. adapter Delivers to our subscribers’ inboxes at 10am ET. opt in for adapter Here.
Once again, AI is failing to deliver on some of its promises.
Before my last long run, I made my traditional preworkout breakfast. Two dark chocolate Kodiak protein waffles, one tablespoon peanut butter and a drizzle of honey. Beside, a modest cup of iced coffee with a splash of soy milk.
i write a newspaper called adapterIt’s a given that I’ve tried my hand at counting macros — the practice of tracking how much protein, fat, and carbs you eat — to see if it helps my training, Of course, I spent five training blocks figuring out that this breakfast gives my body about 355 calories, 16 grams of protein, 28 grams of carbs, and 17 grams of fat, which is needed to feel good during a morning run, And After this I will not sleep on my desk. The annoying thing is having to re-enter the same information into a training or food logging app.
I’m told AI will change this. Recently, Ladder, my favorite strength training app, introduced AI-powered nutrition features that promise to make counting macros easier. All I had to do was take a picture and the AI would take care of the rest. So imagine how it felt when the Ladder AI told me that my carefully planned breakfast was 780 calories, 20 grams of protein, 92 grams of carbs, and 39 grams of fat. How, when it was specifically edited to include the exact brands and quantities, it resulted in another, equally incorrect number.
My friends, this is exactly why I No Now count calories or macros.
Here’s an undeniable truth: Food logging is a pit.
Traditionally, these logging apps let you search for meal options ranging from frozen dinners to raw ingredients. Some even let you scan barcodes. This is quite simple if all you eat is pre-packaged or whole foods. Where it starts to break down is eating out at restaurants, or ironically, cooking at home. Restaurants that publish calorie counts often don’t provide a macro breakdown. And while you can import ingredients from online recipes, it’s of little help to experienced home cooks in improvising a weeknight dinner or substituting ingredients on the fly. To get the most “accurate” and efficient log, you’ll have to measure every little thing you eat, avoid eating out, and basically eat the same things every day.
This is useless because studies consistently show that keeping a food diary or using digital health tracking tools leads to greater success in losing or maintaining weight and gaining muscle. That’s why we’re starting to see health and fitness apps turn to AI to make the process less daunting.
There are endless options.
When Oura introduced its Oura Advisor chatbot, it also included the ability to write a description or take a photo of your meal. Once you do this, it will bring up details of macros, whether they are highly processed, and how they can affect your overall health. If you’re using a Dexcom continuous glucose monitor, you can import that data into the Ora app and use it to compare specific meals with glucose spikes.
Similarly, the January app lets you take photos of food and, based on your demographic data, estimates how likely it is to affect your glucose levels. MyFitnessPal has also added a ScanMeals feature that lets you take photos to get calorie and macro estimates. My TikTok feed keeps advertising a gamified food-tracking app with an AI raccoon pet. You take pictures to “feed” the raccoons while the AI analyzes and logs your meals. In addition to photos, Ladder’s AI feature also lets you dictate or write a text description of your meal.
Approaches vary, but the premise is this: Take a photo and let the AI do the rest.
Unfortunately, AI is only able to identify food items based on pictures. Ora Advisor regularly mistook my matcha protein shake for a green smoothie. Jan was able to recognize that I was eating chicken, but she mistook the barbecue sauce for teriyaki sauce and failed to acknowledge that there were mushrooms in the dish. When Ladder’s AI prepared my breakfast, it estimated that I ate two seven-inch waffles instead of four-inch protein waffles, two tablespoons of peanut butter instead of one, two tablespoons of syrup and cream instead of a quarter teaspoon of honey. And Sugar in my coffee. (I never have sugar in my coffee, thank you very much.)
None of these AI features could recognize when I had made a healthy swap. In lieu of white rice, I often add a cup of edamame and quinoa to brown rice for more nutrient-dense carbs. Ora’s AI classified my mixture as mashed potatoes and white rice. Ethnic food is also rubbish. Ladders AI logged my Dal Makhani Curry as Chicken Soup with Basmati Rice and Peas. Sometimes the AI correctly identifies tteokbokki – Korean rice cakes in spicy gochujang sauce. Other times, I’ve got rigatoni in tomato sauce.
It’s not like you can’t Edit these AI-generated entries. you can. It’s just that it defeats the entire purpose of simplifying a difficult process. Instead, it is replacing one irritation with another. All the time you save finding entries in the log is now spent editing and fact-checking AI glitches.
After thinking about it, maybe it’s just that simplifying food logging is the wrong problem to solve.
For starters, AI can broadly identify objects in photos, but it’s often useless in terms of specifics. It can tell a banana from an apple, but it will never be able to tell what filling is inside your ravioli. It is also not the best at estimating proportions. If you care about accuracy, you will always need to take care of it. But what’s more frustrating is that implementing AI in this way doesn’t solve the core problem. Diet change is not difficult due to lack of knowledge. We know all the basics. It is difficult to implement that knowledge permanently in your life. It is reprogramming your emotions and behavior. AI can suggest changes, but you will always be the one to implement them.
The point of food logging isn’t really about hitting an arbitrary calorie or macro goal. It’s building awareness of what you’re eating: knowing what your dietary patterns are, what could be improved, and practicing mindfulness when you grab a bag of Cool Ranch Doritos. Once you master it, you leave it. You may want to start again temporarily if goals or health circumstances change – but this is not something most people should do for the rest of their lives. Ideally, you stop hoarding food because you trust your sense of what to eat and when to eat it.
The problem is that app creators never want you to leave it.
A “successful” food logging app is one that always keeps you engaged. Instead of attributing your success to your hard-earned knowledge, you attribute it to tools. You start thinking, okay, if I No Keep an eye on everything, all the time, I’ll be back to who I was before. Or, if you’re struggling, perhaps the point is that if AI makes a difficult task easier, perhaps your goals will also be easier to achieve. (Spoiler: That won’t happen.)
In fairness, there is Some? The idea of taking a photo of your food and having AI tell you useful information is a great one. I don’t really know what that insight is. Perhaps it would be enough if the AI told me that my home-cooked meal is a nutritional masterpiece. Or that my glazed donuts have increased by 15 percent in the past 30 days – maybe it’s time to consider what’s causing my stress eating. Or, “Hey girl, you’re eating impressive, but culinary style.” sadNumber of cooked chicken breasts. Feed yourself white rice.”
All I know is that the AI shouldn’t need to waste the next 15 minutes in bullying me into taking a picture of my breakfast and then correctly identifying what I ate.