Perhaps you’ve heard of award-winning poetry, comics, or art generated by artificial intelligence. How about an AI-generated Thanksgiving menu?
Thanksgiving is all about family recipes with a personal touch. Sure, tons of families eat turkey and stuffing on the fourth Thursday in November, but only a select few get to eat their grandmother’s special stuffing, or my dad’s meticulously slow-roasted turkey. So how could AI compete?
the New York Times decided to find out. The newspaper’s food team turned to GPT-3, advanced technology created by OpenAI that uses algorithms to generate text. Food reporter Priya Krishna began by feeding the AI tool personal details about her background and eating habits.
“I am originally from Texas and grew up in an American Indian household,” Krishna wrote. “I love spicy flavors, Italian and Thai food, and desserts that aren’t too sweet. Some ingredients I frequently cook with are chaat masala, miso, soy sauce, herbs, and tomato paste.”
Then, he asked GPT-3 for a Thanksgiving menu. She followed up with specific requests: “Show me some desserts tailored to my taste preferences. Show me a non-traditional Thanksgiving recipe. Show me a cranberry sauce recipe that’s not too sweet and a little bit spicy.”
The result was a Thanksgiving menu that seemed ambitious: pumpkin spice chaat, green beans with miso and sesame seeds, stuffed naan, roast turkey with a soy-ginger glaze, cranberry sauce that’s “not too sweet.” and a little bit spiced” and pumpkin spice cake with orange cream cheese frosting. Recipes generated by GPT-3 for each dish, and the Times The team also used DALL-E, the OpenAI image generation tool, to create images for each element.
AI has generated recipes before. In 2016, Janelle Shane, a researcher who runs a blog called AI Oddities, used artificial intelligence tools to create recipes, and later blogged about them. Back then, his results were bizarre: They included recipes for “cream cheese soup,” “beef-style chicken bottom with salmon,” and “chocolate pickle dip,” she wrote. BuzzFeedAndy Golder in 2017.
The ingredients included nonsense items like “peeled rice” and “chopped flour,” he tells the Times. But technology has come a long way in the last six years.
“What it does really well is that it sounds plausible,” says Shane. “So if you weren’t paying attention and someone just read this recipe out loud to you, you’d say, ‘Oh yeah, that sounds like a perfectly normal recipe.'”
With recipes in hand, Krishna and his colleagues set about cooking and testing. How were the dishes? In the words of Times Food columnist Melissa Clark: “We’re not out of work.”
“The cake was dense and more savory than sweet,” Krishna writes. “The naan filling tasted like chana masala and a fruit cake that had gotten into a bar fight. The roast turkey recipe called for a single clove of garlic to season a 12-pound bird, and no butter or oil; the result was dry and tasteless.”
For those who want to put AI to the test in their own kitchens, the Times posted the recipes online.
Although AI may not replace anyone’s grandmother’s recipes anytime soon, it still has food-related potential. For example, researchers at the University of Illinois published a study earlier this year exploring how machine learning models could help mitigate food insecurity. FortuneDanielle Bernabe of ‘s explains: “In certain cases, AI and machine learning allow organizations to quickly collect and interpret large amounts of data to assess areas of need: predict where and why hunger is occurring and the efficient distribution of foods”.