The Culinary Artistry of Chef Watson

Better home cooks should use AI, not organic ingredients

If you’ve heard of IBM’s AI platform Watson, you probably know it as the supercomputer who won Jeopardy. It turns out Watson isn’t just good at winning game shows; it can also do things like assisting oncologists in making more informed treatment decisions.

A few months ago, I wrote an article titled I’m Conflicted About Blue Apron, wherein I criticized the single-serving, by-mail cooking model driving companies like Blue Apron and HelloFresh. I advocated a more simple and engaged cooking process instead, drawing on whole ingredients made in bulk and recombined as leftovers into an “everlasting” series of meals.

Thinking about that article, I’m glad I discovered IBM’s Chef Watson last week. My affinity for this application of AI can help me explain why my approach to cooking is anything but Luddite.

When I started cooking in restaurants, one chef recommended the book Culinary Artistry by Andrew Dornenburg. She was trying to help me learn the frameworks underlying choices real chefs make. The book was a revolution for me. It contains so many things useful to aspiring chefs: philosophies, menus from famous restaurants, ideas for cooking with the seasons. By far the most useful section, though, is the food match indices — an exhaustive list of foods and their complementary ingredients. Below is an example of the index for apples.

As I became an efficient prep cook and reworked my home cooking practice, I referred to this book often. Conveniently, the most appealing/common combinations are bolded, so your eye is drawn to them. At home, I’d see if my fridge contents fit any listed combinations, and try to put something together without measuring.

Chef Watson serves a function almost identical to Culinary Artistry. Click the LET’S GET COOKING button, and you’re prompted to enter one or more ingredients. The algorithm will then locate complementary foods. Below are listed a series of, yes, suggested recipes — but at least they contain the ingredients selected by you and the computer, which is much better than using recipes that look delicious, but don’t include anything you currently have in the fridge. You can scroll to the right for more suggested ingredients, all listed in order of “synergy” — items that would be bolded in Culinary Artistry are always displayed first.

This is brilliant for home cooks, who haven’t spent a lifetime memorizing (and modifying to individual tastes and beliefs) indices of complementary ingredients, and feel tempted to buy expensive, Instagram-friendly dinners only requiring basic assembly. And it fits my framework perfectly. Watson’s suggestions, like those in Culinary Artistry, advocate home cooking as combinatorics, rather than adherence to a script.

That script isn’t just recipes. It’s a modern aesthetics attached to a wholesome and healthful authenticity. But whether you use organic, locally-sourced ingredients should be up to you. Whole foods are certainly easier to combine and recombine, but Watson allows you to input pre-made foods too, like pancake mix and potato chips.

Interestingly, if you input “apples”, Chef Watson lists some ingredients not mentioned in Culinary Artistry’s list: soy, celeriac, mushroom. On several tries with the same initial search, I also noticed ingredients are displayed in a different order each time. These two facts substitute well for the philosophical discussions at the beginning of Culinary Artistry. Watson subtly reminds me delicious and artful cooking relies on two conclusions. First, we must understand what’s delicious is subjective. Second, we must try many dishes and combinations of ingredients to understand what is satisfying to us — rather than hinging culinary success on outcomes that fit other people’s ideas of what’s tasty, healthy, and in style.

Local & organic.

By the way, if you want to read more about Chef Watson, check out this cool Daily Dot article. Bon Appetit’s article is overly rapturous, but probably also worth a read.