How AI Will Change the Way You Cook

I love cooking so much, I treat it like a sport, and look for excuses to make complicated or unique dishes. I think developing a menu for a bunch of people with differing allergies and preferences sounds like a good time. Until recently, smart cooking tech hasn’t appealed to me, because I assumed it was aimed solely at people who feel clueless enough in the kitchen that they want an assist from a machine.

But over the past year, I’ve interacted with cooking devices—from ovens, to grills, to fridges—that don’t just utilize smart tech, but incorporate AI and machine learning. While AI can’t remove all the labor of food prep or make you love cooking if you don’t, it can make the process of cooking easier by an order of magnitude...not just for novices, but for experienced cooks too. 

AI can help you avoid undercooking (or overcooking) your food

For many people, the dislike of cooking is based in anxiety. It’s hard to screw up a salad, but making anything you can screw up can be intimidating—what if you undercook it and give yourself or others food poisoning? What if you overcook it and destroy an expensive cut of meat? I’ve cooked alongside friends with these common fears—friends who lack my ability to use visual cues to know if a proteins is done, or who have trouble trusting that a little pink is safe.

Tools like a Combustion predictive thermometer can alleviate that anxiety. A smart thermometer probe, the Combustion can be used in almost any situation—a grill, a pot of boiling water, sous vide, the oven or the stovetop. The device has eight sensors along the length of the probe to get measurements of the inside and the outside of whatever you're cooking. Next, AI and a an algorithm are applied to predict when, precisely, you should pull the food off the heat. This means you don’t have to stand over the stove waiting and watching (the app and probe handle that part). It also means that you won’t overcook the food out of fears around food safety, which is something that 50% of people admit to doing.

Combustion specifically altered its algorithm late last year to ensure that food will hit USDA recommended standards, which go beyond simple temperature thresholds. For example, although you commonly think of chicken as being "done" when it measures 165°F, USDA has established you can achieve the same food safety by cooking for a longer time at a lower temperature, as you would using sous vide cooking. The Combustion thermometer can determine whether your food is “safe”, depending on the entire cook history of your protein. This can give the confidence they need to work with protein and, as they see better results, to gain confidence with their cooking. Even as an experienced chef, I love that Combustion does this math for me, so I don’t have to rely on external cues, like how a protein feels to the touch.

There are plenty of other temperature probes with feature's like Combustion's, including the ThermaPro (which i haven’t tested) and the Meater 2 (which I found underwheling).

AI can help cut down on food waste

When I'm grocery shopping, I often forget what is already in my fridge and pantry, and the result is a lot of extraneous purchases—most egregiously when I'm buying fresh foods with short expiration dates. Companies are working to solve this problem. Samsung’s latest fridges incorporate “Food AI,” and use cameras inside your fridge to tell you what you might need to buy more of. Part of their Bespoke line, these fridges come with with AI Family Hub+ and AI Vision Inside. It’s not just that the hub can recognize the fresh foods inside your fridge (up to 33 of them, anyway); it will also offer recipes based on those ingredients.

I haven’t tested the Bespoke yet, but videos of the fridge in action show clear enough imagery that you should be able to easily identify what’s in your fridge from the app, meaning you’ll never have to wonder if you’re out of butter or eggs while you're in the supermarket. 

AI can help you figure out what to make for dinner

Newer technology is taking things farther. AI voice assistants are already embedded in many cooking devices. You can offer the assistant a list of ingredients, or a mood, or a craving, or just allow it to ask you questions, and it will develop meal suggestions for you. 

Even if you don’t have an appliance that can have conversations with you, there are apps aplenty to provide suggestions on the fly. DishGen, MealsAI, and MealPractice all use AI models like Gemini as the underlying engine to produce suggestions based on the language you input, whether that’s a bunch of ingredients or a request based on your mood. 

Using AI while cooking can actually be fun, and save you time

There are a number of “smart” ovens on the market from Tovala, Breville and June, but for the last few months I’ve been using the Brava, an expensive toaster oven with a brain. From a graphical interface on the toaster, you look up any ingredient, and it will generate a list of possible recipes. Choose one, and you'll be guided through inserting a thermometer probe, told where to put the food on the tray and where to put the tray. Then you push a button and walk away. The oven will send you a live video of the food cooking, monitor its progress, and turn off precisely when the food is done.

The oven relies on light technology instead of the normal heating elements you expect in an oven. It focuses heat only where it's needed, for as long as it's needed, specifically to the precise foods you're cooking. Instead of heating up an entire oven, food is cooked from above and below in a very small space. As a result, cooking times are routinely slashed by half, sometimes more. Last week I made sweet potato fries out of raw potatoes. They were perfectly crisped and baked through, in eight minutes.

While the Brava uses only very light AI behind the scenes right now, it’s easy to imagine that in the future, machine learning can help companies process the data coming back from the their appliances to create more recipes and refine the ones that exist, though the tech isn't there yet—I spoke with Brava product manger Zac Selmon, who noted how difficult it is to create a set of parameters to ensure everyone who makes a recipe gets the same results when so many variables involved can differ, from the ingredients, to the environment, to the cook. For that reason, Brava still uses a human team of cooks alongside its data engineers. 

What’s surprised me about the Brava is how much I enjoy not having to think about what I’m cooking. It turns out the tedious part of the process, which involves keeping an eye on a dish as it cooks, is skippable; that the oven that cuts cooking time by half or more, even better: You can enjoy the prep and the results, and not worry about the in between.

I’m excited about the future of AI kitchens

As a self-certified control freak, I shouldn’t like surrendering the cooking or prep process, but it turns out I do. It saves me time and allows me to focus on the parts of cooking I really enjoy. I’ve gifted smart thermometers to a few friends, and the devices have altered mealtime in their homes too. They buy better cuts of meats because they are less afraid of ruining them. They take more risks, and are more confident.

In the future, AI tech will streamline the process more, giving you the ability to manage meal-making from your couch or deck while you spend more time with family and friends. No, a gadget isn’t going to turn you into someone who loves to cook, but it can make cooking a lot more manageable.