With Meghna Chakrabarti
Algorithms influence everything we do now. Who’s creating them? Understanding coders and why how they think is changing how we live.
Want more from the show? You can get messages right from our hosts (and more opportunities to engage with the show) sent directly to your inbox with the On Point newsletter. Subscribe here.
Clive Thompson, journalist who has written about science and technology for the New York Times Magazine, Wired, Smithsonian and more. Author of "Coders: The Making of a New Tribe and the Remaking of the World" and "Smarter Than You Think: How Technology is Changing Our Minds for the Better." (@pomeranian99)
Vaidehi Joshi, writer, speaker, developer and engineer at Tilde Inc. where she works on Skylight, a performance app for software. Creator of basecs and co-host of the Base.cs Podcast, a weekly series that unpacks the basics of computer science and helps foster a supportive community of people learning to code. (@vaidehijoshi)
From The Reading List
Excerpt from "Coders" by Clive Thompson
Chapter 1: The Software Update That Changed Reality
In the early hours of September 5, 2006, Ruchi Sanghvi rewrote the world with a single software update.
A round-faced, outspoken programmer, Sanghvi was 23 years old when she arrived to work at Facebook. Raised in India, she had long dreamed of growing up to work for her father’s company, which lent heavy machinery for the construction of ports, oil refineries, and windmills. But while studying at Carnegie Mellon University, she got intrigued by computer engineering, and then she fell in love with it. It was like constantly solving puzzles: trying to make an algorithm run faster, trying to debug a gnarly piece of code that wasn’t working right. The mental chess colonized her mind, and she found herself pondering coding problems all day long. “You’re at it for hours, you’re not eating, you’re not sleeping; it’s like you can’t stop thinking about it,” she tells me.
Sanghvi was, by programming standards, a late bloomer; she was studying alongside kids, nearly all male, who’d been coding since they were nine and playing video games, and they seemed to effortlessly get it. But she kept grinding away, got good grades, then graduated and got hired for her first job in Manhattan, doing math modeling for a derivatives trading desk.
When she arrived in New York, though, she was horrified by the sight of the gray cubicles at the workplace. She wouldn’t be having much of an impact on the world here. She didn’t want to be a cog in a machine, writing code to support finance work; she hungered to work for a company where the technology itself was the core product, where computer scientists were the main players. She wanted to actually make a product that people used—something tangible, useful. She wanted to do something like Facebook, a site that she’d joined in her last year of college. Now that was an addictive bit of software. She’d log in all the time to stay in touch with college friends who’d recently graduated, checking their pages to see if they’d updated anything.
So Sanghvi bailed on Manhattan, quitting the job even before her first day. She fled to San Francisco, where she got a job at Oracle, the database company. And then, one day a college friend invited her to come by the offices of Facebook itself.
It was a tiny firm, serving only college students; everyday folks weren’t yet allowed to use Facebook. When she walked up to the office on the second floor above a Chinese-food restaurant, she found a passel of mostly young white men, some who’d recently bailed on Harvard: a 21-year-old Mark Zuckerberg walking around in nearly wrecked sandals, Adam D’Angelo (the guy who’d taught a younger Zuckerberg some coding), and Dustin Moskovitz, Zuckerberg’s roommate at Harvard. They worked in a haze of intensity, laptops open on cluttered desks, while playing video games at their nearby dorm-like crash-pad houses, or even while sunning on the roof of the Facebook office. The graffiti artist David Choe was hired around that time to bedeck the walls with murals, one of which depicted “a huge buxom woman with enormous breasts wearing this Mad Max–style costume riding a bull dog” (as early employee Ezra Callahan described it).
They were aggressive about tweaking and changing Facebook, regularly “pushing” new code out to users that would create features like Facebook’s famous “Poke,” or a “Notes” app that let people write longer posts. They were daredevils; sometimes a new feature would have been written so eagerly and hastily that it produced unexpected side effects, which they wouldn’t discover until, whoops, the code was live on the site. So they’d push the code out at midnight and then hold their breath to see whether it crashed Facebook or not. If everything worked, they’d leave; if it caused a catastrophe, they’d frantically try to fix it, often toiling until the early morning, or sometimes just “reverting” back to the old code when they simply couldn’t get the new feature working. As Zuckerberg’s oft-quoted motto went, “Move fast and break things.”
Sanghvi loved it. “It was different, it was vibrant, it was alive,” she says. “People there were like humming along, everyone was really busy, everyone was really into what they were doing . . . the energy was just so tangible.” And as it turns out, Facebook was desperately seeking more coders. It’s hard to imagine now, with the company being such a globe-spanning behemoth, but back in 2005 they had trouble attracting anyone to work there. Most experienced software engineers in Silicon Valley thought Facebook was a fad, one of those bits of web ephemera that enjoys a brief and delirious vogue before becoming unspeakably passé. They had no interest in working there. So Sanghvi arrived in a lucky window of opportunity: young enough to have used Facebook and known how addictive it was, but old enough to have actually graduated college and be looking for a coding job. They hired her a week after her visit, as the company’s first female software engineer.
Soon, she was given a weighty task. Zuckerberg and the other founders had decided that Facebook was too slow and difficult to use. Back in those early days, the only way to know what your friend was doing was to go look at their Facebook page. It required a lot of active forethought. If someone posted a juicy bit of info—a newly ended relationship, a morsel of gossip, a racy profile photo—you might not see it if you forgot to check their page that day. Facebook was, in effect, like living in an apartment building where you had to keep poking your nose in people’s doors to see what was up.
Zuckerberg wanted to streamline things. He’d been carrying around a notebook in which he’d sketched a vision (in his tiny, precise handwriting) for a “News Feed.” When you logged in, the feed would be a single page that listed things friends had posted since you last logged in. It’d be like a form of ESP for your social life. As soon as someone posted an update—Ping!—it would arrive on the periphery of your vision. The News Feed wouldn’t be just a slight cosmetic tweak to Facebook, like a pretty new font or color. It would reconstruct how people paid attention to one another.
And now Sanghvi had to make the News Feed happen. She set to work with a small “pod” of collaborators, including Chris Cox, Matt Cahill, Kang-Xing Jin (known as “KX”), and Andrew “Boz” Bosworth, Zuckerberg’s former teaching assistant at Harvard. For nine months they worked intensely, batting ideas around and then clattering away writing code, while Cox blasted James Brown or Johnny Cash from his laptop. Like the other coders, Sanghvi began programming almost around the clock, staying at Facebook until dawn and then staggering home to San Francisco; after nearly crashing her car from lack of sleep, she moved to a house near Facebook’s office, from which she’d sometimes wander to work in her pajamas. Nobody minded. All the coders blended socializing and working, playing poker or video games at work; during a video interview in 2005, Zuckerberg chatted while toting a red-cupped beer, and an employee did a keg stand.
It was a boys’ club, though for Sanghvi, that wasn’t anything new: The world of computer science she’d known had always been a boys’ club. There were only a few women in her class of 150 at college. She’d learned to yell back when others started yelling, which, in a roomful of cocky young men, was often. Being loud, and a woman, brought repercussions: “Everyone called me super aggressive,” she says. “And that hurt. I don’t think of myself as aggressive.”
But she kept her head down, grinding on the code, because it was mostly what she cared about—and it was thrillingly fun, weird, and hard. Creating the News Feed required her and the other coders to grapple with philosophically hefty questions about friendship, such as What type of news do friends want to know about each other? The feed couldn’t show everything that every single one of your friends did, all day long. If you had 200 friends posting 10 things each, that was 2,000 items, way more than anyone had time to look at. So Sanghvi and the coders had to craft a set of rules to sift through each person’s feed, giving a “weight”—a number that ranked it as more or less important. How would you weight the relationship between two people? they’d ask each other, sitting around the Facebook office late at night. How would you weight the relationship between a person and a photo?
By mid-2006, they had a prototype working. One night Chris Cox sat at home and watched as the first-ever News Feed message blinked into existence: “Mark has added a photo.” (“It was like the Frankenstein moment when the finger moves,” he later joked.) By the end of the summer, the News Feed was working smoothly enough that they were ready to unleash it on the public. Sanghvi wrote a public post— entitled “Facebook Gets a Facelift”—to announce the product to the world. “It updates a personalized list of news stories throughout the day, so you’ll know when Mark adds Britney Spears to his Favorites or when your crush is single again. Now, whenever you log in, you’ll get the latest headlines generated by the activity of your friends and social groups,” Sanghvi explained. The changes, she wrote, would be “quite unlike anything you can find on the web.”
Not long after midnight, Sanghvi and the other coders pushed the update out to the world. News Feed was live; the team cracked open bottles of champagne and hugged each other. It was this type of moment that got her into computers: writing code that changes people’s everyday lives.
There was only one problem: People hated it.
When Sanghvi and the team pushed the code out in those early hours, they clustered around a laptop on her desk to watch the comments from users. She crouched on the ground as Zuckerberg peered down at the screen, clad in a red CBGB’s T-shirt, her colleague KX looming in up high behind Zuckerberg. Everyone was vibrating with excitement. “They were thinking,” Zuckerberg recalled later, “it was going to be good news.”
It was not good news. “This SUCKS” was a typical comment that came scrolling down the screen. Users were in full revolt; many were threatening to leave Facebook or boycott it. Groups had formed with names like “Ruchi Is the Devil.” One student, Ben Parr, had created a Facebook group called “Students against Facebook News Feed” that amassed over 250,000 members in barely a day.
What exactly did they loathe so much? “Very few of us want everyone automatically knowing what we update,” Parr explained. “News Feed is just too creepy, too stalker-esque.” Sure, Facebook had been slow and inefficient before, as Zuckerberg had noted. But Facebook’s users, it seems, had grown to rely on that inefficiency. It gave them a small, pleasant measure of secrecy. They could post a new profile photo, decide it was unattractive, and quickly change it back to the old one a few minutes later, knowing that it was unlikely many of their friends saw the change. But now News Feed was like a pushy, nosy robot that was taking your every post and shouting it to the heavens. Hey, Rita’s no longer going out with Jeff! She’s single again! Check it out!
The coders had been right: Their invention really had changed the way people learned about their social circle. But users weren’t sure they wanted the machinery of their attention upgraded so quickly, and so dramatically.
The uproar grew all day long, and the next day student protestors were camped out in front of the Facebook building, forcing Sanghvi and the other engineers to sneak in and out through the back door. Online, things were even worse. Fully 1 million Facebook users—10 percent of their entire user base—had joined Facebook groups demanding that the News Feed be turned off.
Staff members started arguing about what to do. Two factions emerged, one in favor of shutting down the News Feed, and the other arguing that it was just an adjustment period. Zuckerberg was part of the second camp. Once the initial shock wore off, he believed, the users would discover they liked it. Sanghvi strongly agreed, though she admits part of her insistence in keeping the News Feed was also fueled by engineering pride. “I’d just spent nine months of my life on this, and there was no damn way I was going to get rid of it,” she says.
Zuckerberg’s view won the day. But even so, he admitted they’d moved a bit too hastily and needed to meet their irate users halfway. So the Facebook coders hatched a plan to create some extra privacy settings so Facebook users could prevent sensitive updates from appearing on the News Feed. After 48 hours of pell-mell work, they pushed that privacy code out live. Zuckerberg published an apologetic note publicly on Facebook. “We really messed this one up,” he admitted.
“[We] did a bad job of explaining what the new features were and an even worse job of giving you control of them.” But he was still confident that, in the long run, News Feed would be a hit.
He was right. The feed was unsettling and shocking, but it was also captivating. There was, it turns out, enormous value in seeing a little daily gazette of your friends’ doings. As you checked the feed and saw the status updates scroll by, you could begin to build up a nuanced picture of what was going on in your friends’ lives. Indeed, the day after News Feed emerged, Sanghvi and the team found that people were spending twice as much time on Facebook than before. They were also forming groups much more quickly. It made sense. If you could see that your friend joined a political cause or fan group for a band, you might think, Hey, maybe I should do that, too. Ironically, the whole reason “I Hate News Feed” groups were able to grow so quickly is that they tapped into the power of the feed. (And it wasn’t just silly groups that were forming. In the days after News Feed launched, the second-largest group was one focused on calling attention to genocide in Darfur, and the fourth-biggest was to advocate for breast-cancer research.)
Indeed, you could argue that News Feed eventually became one of the most consequential pieces of computer code written in the last twenty years. Its effects can be seen everywhere, fractally, up and down the patterns of our lives. Facebook users learn that their friends have had babies, see snapshots of their cubicles and vacations; they notice stray jokes and click on cat-meme links. Its massive, shared attention pool has made News Feed one of the surest vectors by which a piece of culture goes viral, from a tear-jerky video of a kind act to an outtake by Beyoncé, from the hopeful, pro-democratic beginnings of the Arab Spring to virulent ISIS recruitment videos. News Feed tied people together and propelled a host of acronymized pop-psychology ailments, from TMI to FOMO.
The feed got people to stare at Facebook a lot—on average, 35 minutes a day for each American. It’s not hard to see why. The feed’s sorting algorithm is designed to give you more of what you like; it pays close attention to everything you do on Facebook—your “likes,” your reposts, your comments—the better to find new items to show you, that, the programmers hope, match neatly with your preferences. Giving people mostly what they want to see makes for a terrific business, of course, which is why Facebook made about $40 billion a year in advertising in 2017. But it turned out that Facebook’s feed, by concentrating everyone’s attention into one funnel, also had some unsettling side effects. It created a central point of failure for civic discourse. If you wanted to seed misinformation, spread rumors, or proselytize hate, the News Feed was a wonderfully efficient tool. By the end of the 2016 US elections and President Trump’s first year in power, journalists discovered that all manner of toxic forces—from white supremacists to merchants of political disinfo clickbait—were gleefully gaming the feed, seeding it with stories designed to whip up political hysteria. Worse, it seemed quite likely that the feed was exacerbating America’s partisan divide, because it was designed to mostly filter out news that didn’t match what you already “liked.”
By February 2017, even Zuckerberg appeared to be wondering what sort of creature he’d electrified into existence. He wrote a 5,700-word note that felt like an oblique and defensive apology for Facebook’s role in today’s political schisms. “Our job at Facebook is to help people make the greatest positive impact while mitigating areas where technology and social media can contribute to divisiveness and isolation,” he wrote.
That’s a curiously cautious mission statement: While mitigating areas where technology and social media can contribute to divisiveness. It’s certainly a more measured rallying cry than “Move fast and break things.” You could read it, perhaps, as a quiet admission that some things ought to be left unbroken.
Adapted from CODERS: The Making of a New Tribe and the Remaking of the World by Clive Thompson, published by Penguin Press, an imprint of Penguin Publishing Group, a division of Penguin Random House LLC. Copyright © 2019 by Clive Thompson.
New York Times Magazine: "The Secret History of Women in Coding" — "Almost 200 years ago, the first person to be what we would now call a coder was, in fact, a woman: Lady Ada Lovelace. As a young mathematician in England in 1833, she met Charles Babbage, an inventor who was struggling to design what he called the Analytical Engine, which would be made of metal gears and able to execute if/then commands and store information in memory. Enthralled, Lovelace grasped the enormous potential of a device like this. A computer that could modify its own instructions and memory could be far more than a rote calculator, she realized. To prove it, Lovelace wrote what is often regarded as the first computer program in history, an algorithm with which the Analytical Engine would calculate the Bernoulli sequence of numbers. (She wasn’t shy about her accomplishments: “That brain of mine is something more than merely mortal; as time will show,” she once wrote.) But Babbage never managed to build his computer, and Lovelace, who died of cancer at 36, never saw her code executed."
Sign up for The New York Times Magazine Newsletter
The best of The New York Times Magazine delivered to your inbox every week, including exclusive feature stories, photography, columns and more.
Washington Post: "Women built the tech industry. Then they were pushed out." — "From harassment allegations at Google to revelations of the biases encoded in artificial intelligence algorithms, Silicon Valley’s sexism has been thrust into the public eye. Just last month, MIT Media Lab revealed that Amazon’s AI facial recognition software has trouble identifying female and darker-skinned faces. In October, Reuters revealed how AI recruiting technology tends to favor male candidates, since it is developed and tested using men’s resumes.
"This all raises a central question: Where are the women?
"Actually, they were initially at the forefront of the industry, back when technologist jobs were considered menial, akin to typists. But as the industry became profitable, male executives developed hiring criteria and workplace cultures that sidelined women. So instead of a space that empowered women, the Internet’s business structures made it a sphere that reinforced masculine biases and patriarchal norms."
The Guardian: "We can teach women to code, but that just creates another problem" — "Technology has a gender problem, as everyone knows.
"The underrepresentation of women in technical fields has spawned legions of TED talks, panels, and women-friendly coding boot camps. I’ve participated in some of these get-women-to-code workshops myself, and I sometimes encourage my students to get involved. Recently, though, I’ve noticed something strange: the women who are so assiduously learning to code seem to be devaluing certain tech roles simply by occupying them.
"Conventional wisdom says that the key to reducing gendered inequality in tech is giving women the skills they need to enter particular roles. But in practice, when more women enter a role, its value seems to go down more."
Madeleine D'Angelo produced this hour for broadcast.
This program aired on March 27, 2019.