Food for Thought
Bernard Bourdain [1] considered himself to be a man of taste.
His refined, sophisticated palate had been honed through countless dégustation and tasting menus. As the eminent restaurant critic of a renown newspaper, he had dined at Michelin-starred establishments helmed by the top chefs in the world. Nothing, however, could have prepared him for Byte, the latest restaurant to be awarded with three Michelin stars—and the very first AI restaurant in the world.
Inside the restaurant lay a sprawling dining room and an open kitchen. Conveyor belts transported a bevy of ingredients, from rosy pink slabs of salmon to fresh green sprigs of mint and coriander. Gargantuan, gleaming robotic arms that moved and spun in an orchestrated dance of smooth, fluid movement, chopping, dicing, slicing, cooking, stirring, assembling and plating in perfect symphony. Flying drones equipped with cameras hovered and darted overhead like a flock of mechanical birds, expediting and inspecting dishes for quality control. And amidst this army of robots and machines, a troop of humans bustled around, from cooks assisting and overseeing their mechanical counterparts to scientists in white lab coats striding around, tasting dishes from pipettes, and scribbling on checklists.
“Welcome, Mr. Bourdain,” A voice said from behind.
Bernard turned, and saw a robot dressed in chef’s whites and a toque hat perched jauntily atop its head. It—he?—extended a robotic hand. “It’s truly a pleasure to meet you in person, Mr. Bourdain. I’m Chef Al, the head chef here at Byte.”
“We’re very honored to have a food critic as esteemed as you to visit our humble establishment.” Chef Al said. “What do you think of our kitchen?”
“It’s—ah—truly impressive. It looks nothing like any other kitchen I’ve seen before.”
“Yes, we pride ourselves as being different here at Byte,” Chef Al said. “You see, I like to think of this more as an experimental laboratory than a conventional culinary kitchen. Here, we employ not just chefs, but also computer scientists, AI engineers, food scientists, biochemists, psychologists, physiologists and more, in our mission to create, craft and cook the perfect meal for our guests, and concoct the ultimate gustatory experience.”
“As you can see, we implement advanced robotics and automation in our kitchen to help with the mundane, manual and menial tasks like prepping and washing, so that we can free up precious time for humans to devote to more important tasks.” Chef Al continued, gesturing grandly around the kitchen with his robotic hands. “We’re a techno-social assemblage of humans working in tandem with technology, pushing the frontiers at the cutting-edge of culinary and technological innovation to advance a new form of avant-garde cuisine!”
Splat.
Chef Al’s grandiloquent soliloquy ended with a veritable splash as one of the giant robotic arms flung an egg across the room, leaving it to land on the wall with a loud, wet squelch. They watched as the egg slowly slid down on the floor in a congealed mess of liquid egg whites, yolk, and broken eggshells. Chef Al winced slightly.
“Sorry you had to see that,” Chef Al apologized, a hint of embarrassment in his robotic voice. “We’re constantly experimenting and innovating with new technologies, but it’s not always successful at first. While our robots are pretty good at basic functions like chopping and cutting, they are still slightly limited in motor functions that require greater mechanical dexterity and finesse.”
“But not to worry, Mr. Bourdain. I can assure you,” Chef Al winked, and reassuringly patted the malfunctioning robotic arm—the same one that had violently hurled the egg across the room (whose remains now lay in shattered smithereens on the floor, being hastily mopped up by a robotic vacuum), and which was now wielding a razor-sharp butcher’s knife as it hacked an enormous piece of meat to tiny, slivered shreds. “You are in safe robotic hands!”
***
Course 1: Just Your Cup of Tea
Bernard was decidedly not assured.
Food is as much an art as it is a science. A magical alchemy and feast of the senses! he thought to himself. How can an AI chef truly capture and understand something as complex and nuanced as taste and smell?
It was enough that fast food chains and food processing companies were manufacturing mass-produced junk that could barely be called food, with their automated factories and robotic assembly lines churning out cheap AI slop consumed by the plebeian masses. But this was fine-dining they were talking about! Were they all doomed to the path of endless AI-fication, where even Michelin-starred restaurants and chefs would be replaced by automation and AI?
Just then, his server, Melissa, arrived at the table, interrupting his thoughts with a steaming cup. “May I present to you your first course, Just Your Cup Of Tea.”
Bernard peered warily into the murky, muddy-colored liquid that vaguely resembled pond water, complete with brownish-green leaves floating on its surface. With trepidation, he took a small sip.
His eyes widened in shock. “What is this?”
To his surprise, it actually tasted… good? The tea was warm, ambrosial, a complex and rich medley of sweet and salty and savory all in one, with a slightly spicy kick at the end.
“This drink was specially mixed and made, as the name suggests, to be just your cup of tea,” Melissa explained. “Our restaurant is fitted with cameras and sensors everywhere to track every move, facial expression and physiological measure of our guests, including their heart rate, body temperature, eye movement and pupil dilation. Chef Al can then analyze and come up with the perfect, personalized beverage to suit your preference, mood, emotion and bodily state.”
“Our thermal sensors picked up that you were feeling slightly cold, with your body temperature being 84.2°F, so Chef Al prepared a hot beverage for you. And because he sensed that you were feeling rather overwhelmed and apprehensive, he picked a blend of warming, comforting spices like clove and nutmeg paired with reishi mushrooms, which not only adds a hint of umami but also has adaptogenic properties for calming and relieving negative stress in the body.”
“And that’s not all. As the meal progresses, we continue to sense and adapt the courses according to how our guests respond to the food. For example, if someone were to take a bite, smile, and have their pupils dilate, then we’ll know that they enjoyed their food. Or if a guest were to frown and have their heart rate spike, then we may infer that they disliked the taste.”
Bernard didn’t know what to say, so he just took another sip of the tea, draining the cup. “I-I’m speechless. This is incredible.”
“No words necessary,” Melissa beamed. “Chef Al can sense it all via the sensors and cameras.”
***
Course 2: Collective Intelligence
Melissa set a new plate down on the table in front of Bernard.
“Your second course is called Collective Intelligence. To create this dish, we drew upon the combined intelligence of chefs from all time, whose memories and knowledge were synthesized and digitized and uploaded into a digital AI consciousness that now resides in Chef Al’s mind. In a way, you could say that Chef Al was taught and trained by all the renowned chefs of the world, including the great Paul Bocuse, Alain Ducasse, Joël Robuchon, Thomas Keller and more.”
Bernard scooped a spoonful, and took a bite. The briny taste of oysters and caviar burst in his mouth (inspired by Keller’s famous Oysters and Pearls?, he guessed), mingling on his tongue with a rich, earthy sauce (reminiscent of notes of truffle from Bocuse’s signature dish, he thought to himself). On the side, there was a quenelle of creamy mashed potatoes (that must be Robuchon's purée de pommes de terre) topped with a chiffonade of herbs.
“It tastes good,” he said, furrowing his brows. “Though rather French, isn’t it?”
Melissa smiled. “The dish is constantly evolving and improving as we add more chefs to the machine learning model. We’re currently in the process of trying to get Gordon Ramsay to sign on, to add some British chefs to the mix.”
“And of course, our Diversity, Equity and Inclusion team is always pushing for more female chefs to be included—but it’s a bit of a struggle, given how few of them there are. We’re negotiating the rights with the estate of the late Julia Child for permission to add her recipes to the training corpus, though her family’s quite reluctant.”
***
Course 3: The Perfect Dish
“Moving onto the main course, The Perfect Dish.” Melissa uncovered the cloche, revealing a beautifully plated dish with a grand flourish.
“This is the pièce de résistance of our restaurant, which took five years to painstakingly develop. Scraping and synthesizing every cookbook and recipe in the world, the computer science team analyzed vast amounts of data and used natural language processing to train Chef Al’s AI algorithm. Our food scientists and biochemists then equipped Chef Al with a dataset of the chemical and aromatic compounds of every ingredient, spice and food in the world, allowing him to isolate and identify flavor networks between ingredients [2]. The machine learning model was then further fine-tuned through reinforcement learning from human feedback (RLHF) from human taste testers.”
So far, the first two courses had surprised him, he would admit. But Bernard was skeptical. This was a bold, audacious claim—could AI really create the ‘perfect’ dish?
He eyed the dish in front of him, which looked suspiciously like a deconstructed pork kebab, and gingerly tucked in. Smoky, tender chunks of grilled pork melted in his mouth; he could taste and smell a pungent, heady mix of aromatic Asian spices like ginger, garlic, onion and chilli; and finally, a fruity, sweet tang of fresh apple and vinegared strawberry compote that provided an unexpected complement and contrast to the fattiness and richness of the meat.
The dish defied his every expectation. “These ingredients are completely unconventional, yet they go surprisingly well together. It’s absolutely sublime; like nothing I’ve ever tasted,” he exclaimed.
Melissa explained: “Chef Al ran a chemical analysis and mapped out a network matrix of all the pairings of aroma compounds in over 10,000 food ingredients, and identified that foods like apple, pork, strawberry share similar flavor compounds, including decanal, a ten-carbon aldehyde often responsible for citrus scents, and gamma-Dodecalactone, a lactone and aroma compound that has sweet, creamy notes found in peaches, apricots and strawberries. This is why the ingredients in this dish, though unexpected, complement one another so well [3].”
“In fact, in our experiments, Chef Al was able to come up with novel flavor-ingredient combinations that were rated as 53% more creative and 73% more delicious than dishes created by human chefs. This dish that you tasted here tonight was the best-performing one in our latest tests.”
“I’d even go so far as to rate this as the most creative and delicious dish I’ve ever had.” Bernard pronounced. “It truly is the perfect dish.”
***
Course 4: Nose and Tongue
For the next course, Bernard was ushered into a different room filled with curious contraptions and machines. Melissa handed him a headset and instructed him to put it on.
“How do I eat with this helmet covering my head?” he asked, perplexed.
“Don’t worry, Mr. Bourdain, you won’t have to do any eating. Not physically, that is.” Melissa reassured him. “We will use biosensors to connect your brain to an e-nose and e-tongue. These two devices are powered by olfactory and gustatory algorithms that will ‘smell’ and ‘taste’ the food on your behalf, then send electrical pulses and signals to activate different neurons in your brain’s cerebral cortex that respond to taste stimuli for different flavors, ranging from salty to sweet to bitter to sour. The headset you’re wearing is also equipped with speakers, which will project AI-generated sounds that are meant to serve as sonic seasoning [4] to complement the dish, making it a multi-sensorial, multimodal experience. Please enjoy, Mr. Bourdain.”
A second later, he heard a soft hiss, and felt a gust of air against his face. His nose was completely blocked by the helmet and he couldn’t smell a thing, yet strangely, he could somehow ‘sense’ the smell of the sea; a briny brackish scent, like salt and seawater. He could taste a meaty, gamey taste that he recognized at once was beef tongue and nose, but infused with the saline taste of the ocean, reminiscent of seaweed. It was the classic surf and turf, but in a completely novel format unlike other dishes he had tasted before. In his ears, he heard the crashing of waves and the song of seagulls. Instantly, he felt as though he was transported to the childhood memory of a summer day at the beach, feeling the warmth of the sun against his skin as the ocean waves drenched his sandy feet.
When he took off the headset, he was surprised to find that his cheeks were wet with tears. He licked his lips, and tasted salt.
***
Course 5: Potato Chip
“Your final dish is a parting gift from me.” When he returned to the main room, he found Chef Al at his table. In his robotic hand, the chef held up a small velvet box. Bernard opened it to find a single, unassuming potato chip.
“A potato chip?” He asked, incredulous.
Chef Al replied, “It may look like a simple potato chip, but this is actually an edible nano-chip. Throughout your dinner today, we recorded your entire experience at a molecular level, extracted the unique chemical signature of the flavors, emotions and sensations you experienced during the meal, and encoded and stored it within an edible data chip. Don’t worry, it’s completely safe to consume, and it will only stay in your body temporarily.”
“Think of it as a memento, allowing you to “re-taste” and savor this exact moment and memory in the future, whether it’s days or decades from now, just by consuming this chip.”
***
The Review
Love at First Byte (★★★★★), reviewed by Bernard Bourdain
Molecular gastronomy is passé—make way for a new wave of algorithmic gastronomy. Helmed by the indomitable Chef Al, Byte is a restaurant like no other, a unique amalgamation of human ingenuity and culinary and technological innovation that offers a bite of something exquisite and exciting.
I had my initial reservations, but it was, indeed, love at first byte, for me.
I used to think that the sensory faculties of smell and taste were solely and exclusively reserved for humans. Byte, and Chef Al, have proven me wrong, showing me that AI can create culinary magic beyond the wildest dreams and imaginations of mankind. Indeed, one day, I fear that even food reviewers and critics like myself will be out of a job too, replaced and made redundant by an AI that can taste, sense, and write better than we do.
I have always considered myself a man of taste—but perhaps, in a future not so far away, there will be machines of taste too.
***
Footnotes
[1] This story is a purely speculative piece of fiction. Any similarity to actual persons, living or dead, is purely (though perhaps not unintentionally) coincidental.
[2] In 2011, by analyzing chemical and aroma compounds found in food, researchers were able to create a network to explore the impact of flavor compounds on ingredient combinations (Ahn et al., 2011).
[3] This was inspired by a real-life recipe by Chef Watson, an Artificial Intelligence model created by IBM in collaboration with Bon Appetit, that analyzed recipes and came up with creative dishes with surprising flavor and ingredient combinations, including an unexpectedly tasty Vietnamese apple kebab (Billow, 2014).
[4] This dish is inspired by Heston Blumenthal’s (a British three Michelin starred chef known for his innovative modernist gastronomy) Sound of the Sea, a dish that makes use of a technique he termed as ‘sonic seasoning’, pairing food with sounds of seagulls and ocean waves.
***
Kai Xin Soh is a Ph.D. student in the Media, Technology, and Society program at Northwestern University. As an academic and artist, she explores the impacts and implications of Artificial Intelligence on the creative arts and industries through a socio-cultural, ethical, and critical lens. In her free time, she enjoys creative writing and creating art, whether it's painting with watercolors or experimenting with generative art made with Stable Diffusion. X: https://x.com/kaixinsoh