A.I. In the Kitchen

It might surprise you, but we’ve had ChatGPT and other Artificial Intelligence (A.I.) models for three years now. I know a major concern when it was first released was job security and the business implications of the software. Thankfully, the employment landscape hasn’t drastically changed due to A.I.—yet. As we continue to integrate A.I. into our lives, it’s important to maintain a balance between sourcing your own information and leveraging artificial intelligence. Below are some of my thoughts on the future of A.I., specifically in the kitchen.

When Artificial Intelligence first came on the scene, I originally thought the kitchen would be immune to its automation. How could a system reduce all the variables of cooking into a series of inputs and outputs? The simple answer: it can’t, but it can. It all depends on what you're eating.

Any major drive-thru fast food chain could easily implement a kitchen run by A.I. The food has become so standardized and processed that the only things missing are the specific equipment and the software to communicate between them. Unfortunately, A.I. solves the communication issue, and specialized equipment is manufacturable with the right budget. Most fast food is already cooked in an assembly line fashion—with specific temperatures, timers, and a depressingly low level of human involvement. How different would it be if a machine put the food in, then took it out when the timer went off? Not much. This isn’t the type of food or cooking I celebrate, and it certainly goes against Loqavore’s mission, but the fact that A.I. could run a fast food restaurant really illustrates why we should be cautious about how we use it. Outside of the speculative future of A.I. in fast food, I don’t think it can compete with made-to-order restaurants.

The beauty of made-to-order restaurants is that there are too many variables in cooking to reduce them all to simple inputs and outputs. No piece of protein, whether it’s a cut of steak, a fish, or a chicken breast, is exactly the same. Produce can vary wildly from item to item; sometimes even one piece has inconsistencies within itself. That’s still not accounting for seasoning, temperature, managing heat, or most importantly, taste. Those small differences between ingredients have a cascading effect throughout the dish. If you're pairing a protein with something acidic, and some cuts have different fat content, you'll have to adjust the acid to taste. Consistency in the kitchen is driven by taste, not by the recipe. Recipes are vital and offer the bones of a dish, but our ability to taste lets us adjust for the natural variation in ingredients that will never be exactly the same twice.

Artificial intelligence can’t bridge that gap. And I’m not sure it ever will. There’s a subtle beauty in human intuition that I think plays a bigger role than we realize.