A person falling into a giant blue flower
Glenn Harvey

The Next Olympics Mascot Might Have Been a Mutant Morning Glory

Scientists dreamed of genetically engineering a flower patterned in the Games’ blue-and-white checkerboard emblem.

The 12th-floor apartment of one of Long Island City’s waterfront towers features both spectacular views of Manhattan and a small yet state-of-the-art bioengineering lab, tucked into the spare bedroom. Sebastian Cocioba, a 29-year-old college dropout and self-styled “plant hacker,” has lived there with his parents for the past decade. And, for the past three years, the condo has also been home to a top secret, gloriously quixotic enterprise: the project to genetically engineer a flower that would serve as the official mascot for the 2020 Tokyo Olympic Games.

When I first visited, on a bright-blue morning in January, Cocioba led me into the kitchen. There, he used the baked-potato setting on the microwave to warm a flask filled with a gelatinous goop of agar, sugar, and fertilizer. Once heated, the mixture loosened up into a free-flowing, straw-colored liquid that smelled, coincidentally, of potato. Meanwhile, Cocioba opened the fridge, reaching into a drawer divided in half—deli meats and cheese on the left, hotel-shampoo-size bottles of chemicals on the right—to retrieve two vials of plant hormones. “Look, we eat these in pretty decent amounts in salad,” Cocioba said, in response to my raised eyebrows. “My parents have kind of gotten used to the whole concept of this by now.”

Down the hall, in the lab, Cocioba assumed the role of patient tutor, while I switched on the laminar flow hood, gloved up, and used a pipette to transfer each hormone, in a carefully measured ratio, to the agar jelly. My task for the day was to insert a small genetic sequence into a white petunia—a small but important step toward the larger goal. Our tool was a plant pathogen known as Agrobacterium tumefaciens, which hijacks its hosts by sending out small packages of membrane-wrapped DNA capable of inserting themselves into the other plant’s genome.

Cocioba and I prepared the petunia for infection: he by stripping off some leaves the day before and leaving them to sterilize overnight in a weak bleach solution, I by using a hole punch to cut out dozens of neat circles of leaf tissue that I then tweezered gently into petri dishes filled with our cooled, yellow jelly. The freshly injured leaves emitted a chemical distress signal that was undetectable to me but that Cocioba assured me would act as a red rag to the bullish Agrobacterium.

The genetic sequence we were hoping to infect them with was a probe, capable of finding and binding to a target sequence in the petunia’s DNA, and it held a tail of green fluorescent protein that would only unfold enough to glow once a successful bond had formed. Cocioba had ordered the probe online and stored the vial, containing a single clear droplet filled with enough genetic material for 50-odd experiments, in the freezer, beneath a bottle of vodka and some tater tots, until he was ready to add it to our batch of Agrobacterium.

Dosing the plant tissue with hormones was a warm-up for the main event: Together, the chemicals would return chunks of adult plant—tissue that had already become root, stalk, or leaf—back to an embryonic state. Post-infection, Cocioba would use the same hormones, in different ratios, to organize the cells in each proto-plant disc back into the constituent parts of a seedling that he could cultivate, and that, if our experiment was a success, would emit an eerie greenish glow under a fluorescent microscope.

In October 2015, the interaction designer Kevin Slavin was in Tokyo, meeting with senior executives at Mori Building Company, Japan’s largest commercial landlord. Slavin is a skinny geek who trained as an artist; developed one of the first location-based phone games, Pokémon Go’s predecessor; and then founded the Playful Systems lab at MIT’s Media Lab. He was in town to present the results of a successful collaboration with Mori that had used bees housed atop the company’s properties to map the city’s microbiome. (The bees functioned as distributed surface-sampling devices, and by collecting their waste from the hive every week and sequencing the DNA found within it, Slavin was able to conduct a microbial census on a neighborhood-by-neighborhood basis.)

The conversation turned to the future. Masa Ogasawara, Mori’s sphinxlike executive managing officer, asked whether Slavin had any ideas for a project for the Tokyo Olympics.

Slavin, who had spent a considerable amount of time in Tokyo for the bee project, had noticed the city’s preparations, and found them vaguely depressing. “I actually love the Olympics,” he told me. “But I love what they are intended to be, and I really respond badly to the crass commercial qualities of it.” Meanwhile, to sequence the microbial DNA collected by his bees, he had also been spending time with the computational biologist Elizabeth Hénaff, and as he began to learn about new gene-editing techniques such as CRISPR, he realized that engineering life was no longer science fiction—it was the imminent future.

As he reflected on what he disliked about the Olympics—the tchotchke-choked monetization that accompanies an otherwise stirring display of human effort, teamwork, and excellence—Slavin wondered what its opposite would be. A true Olympic mascot, he felt, should be “a source of delight and wonder and beauty, and actually add something to the planet instead of just ending up in a landfill somewhere.”

Slavin imagined designing a new form of life, to be collectively grown and given away—perhaps a tree, genetically modified so that its leaves expressed Olympic colors. He told Ogasawara that he had an idea, but that there was no way Mori would be bold enough to do it. This, unsurprisingly, was like catnip to the powerful executive, and the company quickly signed on to support the creation of the world’s first genetically modified Olympic mascot.

When Slavin got back to New York and described his vision to actual biologists, including Hénaff, they gently pointed out that any plan that involved growing a tree from an embryo in five years, let alone engineering an entirely new variety and then propagating it, was hopelessly ambitious. A genetically modified flower, on the other hand—well, that might just work.

The first Olympic mascot was Waldi, a striped cartoon dachshund who made his debut in Munich, at the Summer Games of 1972. Designed by Otl Aicher, better known for the Lufthansa logo, it is also the most tasteful mascot to date. London’s 2012 one-eyed Mr. Blobby lookalike, Wenlock, is probably the field’s nadir, but the brief to represent the host country’s cultural heritage in a “festive” way and appeal to a younger audience has rarely resulted in design excellence. “The problem is that all these things are done by consensus,” says Paola Antonelli, a senior curator of architecture and design at New York City’s Museum of Modern Art. “Did you see the overweight bald eagle from Los Angeles?”

The Tokyo Games, by contrast, were off to an aesthetically pleasing start. The official logo, unveiled in April 2016, consisted of 45 dark-blue rhomboids arranged into a wreath. Officially named “Harmonized Checkered Emblem,” it is a minimal masterpiece designed by the artist Asao Tokolo, who uses a ruler and compass to create repeating patterns. In his studio in Tokyo, Tokolo sketched a quick diagram to show me how the logo’s color was derived from the angles of three rectangular forms—100, 86, and 50, which, when translated to the cyan, magenta, and black of a printing press, produce the same deep indigo traditionally worn by samurai.

This checkerboard pattern is called ichimatsu in Japanese, after an Edo-era kabuki heartthrob, Sanogawa Ichimatsu, who habitually performed in a patterned costume. Mathematicians have calculated that the logo’s rhomboids can be rearranged into half a million new patterns, Tokolo said. He showed me a printout of a paper analyzing his logo, titled “On the Enumeration of Chequered Tilings in Polygons.” This combination of rule-bound repetition and near-infinite variation makes the logo into a universal code, he told me, opening his laptop to play “Pachelbel’s Canon” as an illustration. “This way,” he said, “it’s shareable, transferrable, and transformable”—like music, math, and, I couldn’t help but think, DNA.

As soon as he saw Tokolo’s logo, in early 2016, Slavin knew that the Olympic-mascot flower should be engineered to have an indigo-and-white ichimatsu pattern on its petals. At the same time, Hénaff suggested using the morning glory, a flower she loves—she has a tattoo of a purple morning-glory vine covering the entire side of her body—and that she knew, from time spent working in Japan, held a particular significance in the country. “You see it growing in little postage-stamp gardens in the older neighborhoods of Tokyo,” she told me. “And then you realize there’s morning glories everywhere, in illustrations and artwork and all the lovely printed fabrics.”

In much of the rest of the world, the twining vine is seen as a weed, even a nuisance—its touch-sensitive tendrils help it climb walls and facades, hooking into tiny cracks and turning them into fissures. But in Japan, asagao, which translates to “morning face,” is a cultural icon, its imperial-blue, trumpet-shaped flowers symbolizing high summer in the same way that cherry blossoms signify the arrival of spring.

The species is thought to be native to Central America, where the psychoactive alkaloids found in the seeds of some varieties were used in Aztec rituals, but according to Reiji Iwabuchi, a scholar who has curated a series of exhibitions on morning glories at the National Museum of Japanese History, the flowers were brought to Japan from China in the ninth century. The earliest Japanese mentions of the plant cite its usefulness as a laxative, he told me, and it is pictured in a set of scrolls from 1164, preserved at the Itsukushima Shrine.

In his dark, book-lined office on the campus of Gakushuin University, just steps from the neon excess of Shinjuku in Tokyo, Iwabuchi showed me a series of reproductions illustrating the next phase in the morning glory’s rise to popularity. By the early 1700s, Japan’s doors had been closed to the world for nearly a century. In Tokyo, then known as Edo, but already one of the largest cities in the world, culture flourished and a distinctly Japanese relationship with nature, as well as the craft of expressing its essence in miniature, was refined. Bonsai trees became popular, as did suiseki, or the art of selecting and displaying stones that represent larger landscapes, such as mountains, canyons, or coastlines. Flower vendors walked the streets, selling chrysanthemums and camelias, while feudal lords rewarded their favored retainers with potted plants.

Iwabuchi pulled out a print of a gorgeous gilded screen from the mid-18th century. The artist, Jakuchu Ito, is known for his depictions of chickens, and this print showed a rooster perched on one leg, head turned to face his own dazzling black-and-white tail feathers. In the background are a handful of sunflowers and, woven through them, a spatter of morning glories. Instead of the standard solid blue-purple, the flowers are as variegated as the rooster’s own plumage—there are solid white flowers, but also white flowers speckled with blue, or sporting a series of blue wedges of different sizes. This, Iwabuchi told me, is the first record of a floral phenomenon that was soon to sweep the city: the cultivation of henka asagao, or mutant morning glories.

The first morning-glory craze lasted 30 years, beginning in 1800, and infected all levels of society. The trend was for differently colored and patterned flowers—speckled, striped, albino, half-and-half, pink, maroon, and even a creamy-yellow “phantom” morning glory that modern breeders are still unable to reproduce. Monks and samurai raised thousands of morning glories in their gardens as a side hustle, selling regular purplish-blue flowers to the common people and sought-after mutants to wealthy collectors. Anthologies cataloged the varieties, naming each mutation after literary characters; woodcuts depicted shoppers carrying potted morning glories back from shrines, as well as morning-glory viewing parties in late summer. One kabuki actor, who dressed in a morning-glory print and went by the stage name Asagao Senbei, or “Morning Glory Rice Cracker,” had an entire routine that involved starting vigorous fights, then quickly fading and losing, in the same way that a morning glory blooms at dawn, only to shrivel up by the time lunch is over. But then, Iwabuchi said, for reasons that remain hard to discern, public interest shifted, and the Japanese sacred lily and painted fern were suddenly all the rage.

In the 1850s, another chance mutation birthed a second, mini-boom: This time, growers competed to produce morning glories with curling, ribbonlike petals as opposed to the standard trumpet, or with leaves that forked like a snake’s tongue. “Mutant morning glories became a status symbol,” Iwabuchi explained, “and flower nerds—morning-glory maniacs—competed between themselves to select and maintain the most spectacular deformations.”

The final wave of morning-glory popularity came in the 1870s, after the arrival of American gunboats had forced the country to open up. In a wave of nostalgic, nationalistic sentiment, social clubs devoted to raising mutant morning glories formed, keeping many rare strains alive. Iwabuchi showed me a black-and-white photo from 1910, showing what he called “a nerd garden”—thousands of seedlings, growing in pots under netting in preparation for the annual morning-glory fair in the Tokyo suburb of Iriya.

Since his first exhibition on the subject, 15 years ago, Iwabuchi has seen a renewal of interest in the flower. Japanese schoolchildren grow the basic morning glory as a summer project in elementary school, making it a piece of nature that all the country’s citizens have some connection to, but mutant sales have recently become a major source of revenue for the museum. “I believe we are at the start of the fourth boom,” he said.

At this point, the idea of a mutant morning glory, engineered to express the recombinant code of the official Olympic logo, began to assume an almost unbearable rightness. It blended Japan’s unique culture and history with the latest technology. The fleeting symbolism of the flower even embodied something of the magic of the Olympics—an event that briefly captures the world’s attention every four years, before disappearing again.

Meanwhile, those involved in the project found something that spoke to their own, deeper desires. For Slavin, it was this idea that designers could make a living organism rather than a product, something that would add to the world rather than extract more of its resources. For Hénaff, the project’s almost whimsical goal created an opportunity for a more meaningful conversation about the rights and wrongs of genetic engineering—a less fraught way to think through what it means for humans to have user-level privileges over other species’ genetic identity. In a world where proponents of genetically modified organisms say they are needed to solve world hunger, and their critics say they are being used by corporations to perpetuate inequality, she explained, “oftentimes, the conversation about the technique and the conversation about the goals get kind of muddied.”

For Mori, the project captured something of the subtle Japanese conception of the relationship between humans and nature. Japan is the largest industrialized country that still actively practices an indigenous, animist religion, in which nature does not belong to humans, but vice versa. “We have always designed life,” Jun Fujiwara, Mori’s director of special projects, told me, pointing out that the foods and drinks most central to Japan’s identity—sake, miso, natto—rely on microbial communities that are a hybrid of nature and culture. From this perspective, perhaps a genetically modified morning-glory Olympic mascot could be a kind of Shinto GMO—an organism that embodied not only deep respect for the astonishing ingenuity and beauty of the natural world, but also a sense of necessary awe at the craft with which humans can shape it.

And, finally, for Sebastian Cocioba, the plant hacker that Hénaff brought in to work on the project full-time, the morning glory offered a step toward his life goal of becoming a flower designer. As a teenager, Cocioba funded his studies by flipping Home Depot orchids: When his local store threw out plants that had ceased to flower, Cocioba retrieved them, dosed them with blue light and hormones, and sold them back. He built his lab by buying broken equipment on eBay and fixing it. This project promised a salary and a chance to make something that would be seen on the world stage.

In early 2016, with the flower and pattern decided and Mori funding secured, the team—Slavin, Hénaff, and Cocioba—gathered around Cocioba’s parents’ dining table to brainstorm. In their excitement, they used a blue whiteboard marker to scrawl diagrams, sketches, and chemical formulas all over the white kitchen cabinets, where the drawings remain to this day.

As Hénaff and Cocioba broke it down for Slavin, the flower presented two distinct challenges: making a white morning glory with the ability to produce indigo blue, and then manipulating the expression of that color over time. As it turns out, true blue is actually quite rare in flowers, for evolutionary reasons: Pigments initially evolved to protect organisms by absorbing ultraviolet light, and tweaking those metabolic pathways to reflect more blue and absorb reds at the other end of the spectrum presents an almost insurmountable biochemical challenge. Most “blue” flowers are purple-tinged, and even hydrangeas and cornflowers cannot achieve a true blue without some additional help from acidic or magnesium-rich soil.

A blue-and-white checkerboard-patterned flower
Glenn Harvey

What’s more, the genetically modified version of the multistep process by which plants naturally transform pigments from red to purplish-blue has been patented by Suntory, a Japanese whiskey company. In 2004, Suntory partnered with Florigene, an Australian genetic-engineering firm, to create what it claimed was the world’s first true-blue rose. “It drives me nuts, because it’s purple,” Cocioba told me. (The rose technically contains blue pigments, but appears indisputably lavender-colored in press images.) Cocioba’s proposed solution was to avoid the nine genes required to produce purplish-blue in plants altogether. Instead, he planned to engineer a white morning glory into which he would insert a single gene borrowed from coral, where it would express an intense ultramarine protein.

Typically, a petal’s cells grow in an unbroken line outward, which is why two-tone flowers usually feature either landing strips or radial, halo effects—the color that each cell expresses at the center is the color it will express along its entire arc of growth. To create a checkerboard pattern, Cocioba and Hénaff would have to engineer a switch of some sort, so that they could flick the blue protein on and off at regular intervals within the four-hour window during which morning-glory petals develop from bud to flower.

Luckily, the morning glory had already mastered that trick on its own: Think of the blue-speckled white flowers in the background of Jakuchu Ito’s rooster painting. Such mutations are due to transposons, mobile DNA sequences capable of hopping into, say, the genes that synthesize blue pigment, temporarily disrupting their ability to function and creating a white patch, before jumping out again, at which point the color returns. Every living thing has these “jumping genes,” but they are particularly active in morning glories, where scientists theorize that selection, both by pollinators and, latterly, humans, has favored the resulting diversity.

Hénaff’s earlier research had focused on transposable elements, and on Cocioba’s parents’ freezer drawer, she’d illustrated the idea of a built-in on/off mechanism using a series of arrows and squiggles. “The idea was, ‘Well, you know, if this switch already exists, then maybe we can just tweak its activity and decide when it goes on and off,’” she said. Not all the natural triggers for a gene jump are known, but some are well described, including an insect attack, water stress, and temperature change. Standing in front of the fridge, pen in hand, Hénaff suggested putting their prototype plants in an oven and setting the thermostat to oscillate as the flower developed. That way, they could figure out the exact pigment on/off sequence needed to make a checkerboard, “and then find a way to genetically hard code that timing into the final flower,” Hénaff explained.

The science seemed ambitious but not impossible. Hénaff did a literature review, and saw that morning glories had been successfully engineered using Agrobacterium and cultured in a lab. She even found descriptions of engineered pigment transformations, although none included the rapid on/off oscillation this design would require. She consulted with other biologists, who agreed: feasible, potentially even in a quick time frame. Slavin emailed a photo of his kitchen-cabinet flower sketch to a friend and asked him to render it in Photoshop, then built a Keynote presentation. Cocioba ordered DNA, reagents, and a packet of imperial-blue morning glory seeds from Japan. Through an intermediary, Mori helped set up a series of meetings with representatives from the Office for the Promotion of the Tokyo 2020 Olympic and Paralympic Games, as well as other government officials.

The enthusiasm for the flower was palpable. Slavin, who does not speak Japanese, remembers one man turning to the guy next to him and muttering something that included the words fast money. “They were like, ‘We understand that this is speculative and we respect that,’” Slavin said. “‘We know you might not make it. So we’ll give you a year. Is that fair?’ And we’re like, ‘Yep, totally fair, totally doable.’”

A few months in, the team confronted an issue. Cocioba had never worked with morning glories, so his first step was just to get it to grow in tissue culture, like our hole-punched petunia leaves. “For the first two months, I did all of the things,” he said. “I threw the entire book at it, trying to get it to regenerate—and no dice.”

Hénaff went back to the papers that described morning-glory cultivation in tissue culture and discovered a note explaining that the morning glory was missing the one gene that would allow its leaves and stem to return to an embryonic state in response to plant hormones. Instead, the authors wrote, the only morning-glory tissue that will regenerate in culture is an actual immature embryo. Suddenly, Cocioba had an urgent need for seed.

“So I started growing morning glory like it was weed,” Cocioba said. He rented an artist’s studio a few blocks from his parents’ condo in Long Island City, bought grow lamps and metal racks from IKEA, and threw all the chemicals he could at the baby plants to get them to flower fast. And then, nearly three months later, once they had flowered, Cocioba had to figure out how to pollinate his morning glories so that they would set seed. “I found a company online that sells dead bees on sticks,” he said. “They work wonderfully, but let me tell you, manually pollinating thousands of flowers took for...ever.”

When Cocioba finally harvested a seedpod, it took him two hours of surgery to get an embryo out. “It was half a millimeter long,” he said, showing me a photo on his phone of a greenish dot. “If you pinch it, it dies; if you look at it funny, it dies; if a butterfly farts in Russia, it dies.”

The morning glory was not the only recalcitrant player. Later that year, Kevin Slavin left his faculty position at the Media Lab, and the accounts-payable department at MIT proved reluctant to funnel Mori’s funds to a 20-something in Long Island City. Back in Tokyo, Mori was handling the project as it might manage any business initiative: with an insistence on timelines, check-ins, reports, and deliverables. “It was planned like a construction project and funded like a government project, meaning the money came in a year later,” Cocioba said—leaving him wrestling with an uncooperative flower, working on other projects to pay bills, and missing deadlines left and right.

He and Hénaff made an executive decision: They would prototype using petunia, which grows happily in culture and flowers faster. To try to speed things up, Hénaff spent weeks sequencing a white and a purple petunia and assembling their genomes, so that Cocioba would have an accurate road map to work with. “It’s borderline biblical, the level of precision that we have on these two plants,” Cocioba said.

In spring 2017, with no flower in sight, the Japanese Olympic Committee announced a mascot design competition. Schoolchildren were invited to vote for their favorites, in a bid to make the Tokyo Games seem participatory and transparent following repeated accusations of graft. The resulting mascot, Miraitowa, is a big-eyed cartoon character that looks like a blue-and-white-checked cartoon kitten. According to the mascot-selection panel, he is “imbued with energy that will cheer up and cheer on the athletes,” although critics have described him as a “Pokémon refugee.” Just as Slavin had hoped to avoid, factories in China are already churning out Miraitowa plushies by the million. Miraitowa T-shirts, baseball caps, collectible pins, and stuffed dolls are expected to bring in about 14 billion yen in revenue for the Games’ organizers.

With the mascot decided, Takeo Hirata, head of the Office for the Promotion of the Tokyo 2020 Olympic and Paralympic Games, told Slavin that his genetically modified morning glory could instead become an Olympic emblem—a new category of Olympic symbol. Hirata suggested that the checkerboard flower could replace posters and banners in some places, and that the Japanese government could still distribute it to all public primary schools in spring 2020. Which meant that the pressure from Mori continued, with the company sending Jun Fujiwara to New York City to visit Cocioba’s apartment lab on a regular basis.

By now, Slavin was distracted by the demands of a new job. Meanwhile, after finding that, for months, Fujiwara had removed her name from reports she’d prepared and left her off update emails, Hénaff had distanced herself from the project. The final straw came when Mori arranged to bring Cocioba and Chris Mason, the Cornell geneticist in whose lab Hénaff had been a postdoctoral student when the project began, to Tokyo to present the group’s research. Though both Cocioba and Mason considered Hénaff the project’s principal scientist, she didn’t hear about the trip until Cocioba called her to see where she was staying. “I wasn’t even contacted about it,” she said. At that point, disgusted and demoralized, she was done. (Fujiwara, when asked about his reasoning, blamed limitations of time and space.)

In October 2017, after their presentation in Tokyo, Mason and Cocioba were drinking beer at the Park Hyatt bar, better known as the setting for a scene in the movie Lost in Translation, when Mason received an email from Fujiwara requesting that he present an update on the flower project to Mori’s president the next day.

Until that point, Mason’s involvement in the project had largely been limited to providing access to computing power, as well as managing funds following Slavin’s departure from the Media Lab. Scrambling, Mason did something no one on the team had done before: a Google Image search for checkerboard-pattern flower. Astonishingly, one came up—the snake’s head fritillary. Cocioba then entered the plant’s scientific name, Fritillaria meleagris, into a scientific database and told Mason that it has one of the largest genomes on the planet, at about 156 billion base pairs to the human genome’s paltry 3 billion.

“As soon as I told Chris that, his eyes lit up, because he loves superlatives,” Cocioba said. “He’s like, now I have a purpose—we’ll sequence it!” A chunk of the Mason Lab’s funding comes from NASA, and somehow, over the course of that evening, Mason—a creative thinker who has also written a short volume of genetic poetry—decided that the process that produced a checkerboard pattern in the snake’s head fritillary could serve as a model for the mechanisms that cause changes to an astronaut’s DNA in space.

The next day, Mason’s PowerPoint contained NASA’s logo and a photo of the flower, grabbed from the Brooklyn Botanic Garden’s website. Slavin received a copy in an email update from Mason, in which the project’s goal is described as “astronaut protection,” and Slavin is listed simply as an adviser. “This is out of nowhere,” Slavin recalled. “NASA? What? I’m like, How far up the river has this been sold, by how many people?

A few hours’ west of Tokyo by bullet train lies the Japanese National Institute for Basic Biology. When I visited, one morning in March, a slight, baby-faced scientist named Atsushi Hoshino loaned me a pair of orange Crocs and led me into his greenhouse. Inside, spilling over wire racks and tangling together, was a jungle of mutant morning glories—nearly 200 in total, he told me, each one unique. An enormous deep-burgundy flower with a white center almost sparkled, its petals velvety under the lamps; in the corner were white flowers, each with one or two wedges of pink, and on the rack above, deep-blue flowers with rays of purple.

“I screen 4,000 baby plants each year to discover new mutants,” he told me. The normal morning glories end up in the incinerator; the mutants go into the greenhouse. In his lab, next door, 800 seedlings were growing in trays under lights, each just a couple of inches tall and too young to flower. Hoshino showed me his only hope from this latest batch; its sole leaf was split down the middle, half lime green and half emerald. “A transposon has disrupted the gene for chlorophyll,” he explained, with a shy smile.

Mutant morning glories have been at the center of Japanese genetic research for more than a century, Hoshino told me. The 19th-century samurai and monks who competed to breed the most spectacular new mutations had no idea what a gene was, let alone one that jumped. But morning glories were the topic of the first bulletin of the Japan Breeders’ Association, in 1916, and after the Second World War, the Japanese government collected all the mutants it could find from hobbyists and founded a national research center. Since then, by analyzing these mutants, Japanese researchers have discovered many of the genes that determine flower and leaf shape, color, and pattern.

As a doctoral student, Hoshino wanted to study rice, in order to develop useful improvements for Japanese farmers. “In my lab, there was a better student, and she chose the rice project,” he told me. “So my boss gave me mutant morning glories for my thesis topic.” Decades later, Hoshino feels that he was the lucky one. “There are hundreds of possibilities for patterns,” he said. “We only fully understand a few.” He pointed to a blue flower with a white halo at its edge. “Why is the silencing expressed only at the margin?” he asked. “No one knows.” Although he used to think of genes as stable, heritable sets of instructions, morning glory has taught him to see them as a dynamic system, whose shifts we have yet to comprehend. “We need mutant morning glories to teach us how this works,” he said.

Back in New York City, Cocioba unknowingly echoed Hoshino’s words. “In the end, morning glory taught us that we know nothing about morning glory,” he said. In October 2018, in an abrupt email, Mori pulled the plug on the project, but Cocioba continued working on his petunia prototype on evenings and weekends. The experiment we did together was a test of his genetic guide; it was the final step before he attempted to insert his new blue gene from coral. With Mori out of the picture, Hénaff was once again involved, and together they were looking for funding.

The morning-glory mascot had taught Hénaff something, too: how not to manage a project. “Really, the tragic flaw was the hubris of thinking that we could deliver a new plant on a human schedule,” she told me. Already, she said, the project had yielded enough in the way of new knowledge for at least a couple of papers. If, in the future, Cocioba succeeds in turning his white petunia blue, the economic reward could be substantial: Experts have estimated the value of a true-blue rose at many millions of dollars. If they manage to design a switch to turn the gene producing that color on and off during the bud’s development, the impact will go far beyond horticulture—although Cocioba warned me that, once humans have the ability to create petal patterns on demand, we should expect to see Nike swooshes and McDonald’s arches on our flowers shortly thereafter.

The morning-glory mascot project was, unquestionably, a failure. Kevin Slavin’s ambitious vision of a beautifully designed plant, whose seeds could be given away to visitors and raised in pots by schoolchildren on the balconies and in the alleyways of Tokyo, will not become reality—if it ever could have, given the constraints on releasing gene-edited organisms. “From a cultural perspective, it was a clusterfuck,” Slavin concluded. “From a scientific perspective, it may yield something that’s more important than anything we could have ever imagined.”

And then Slavin told me another story, about the poet John Giorno. In 1970, Giorno worked with MoMA to create an installation called “Dial-a-Poem,” in which visitors were invited to call a number to hear one of 50 poems. “They had to figure out how to do it—they had to hook up tape recorders and solve a bunch of difficult problems,” Slavin said. “And they did, and that idea led to a multibillion-dollar industry of 1-800 numbers.”

Slavin’s point was that when designers and artists flirt with science and technology, they can unlock new ways of thinking, from which new ways of being and doing emerge. In the case of “Dial-a-Poem,” the unexpected consequences were rather banal, Slavin said. “But, in this case, it might be something great.”