I’ve known Moran Cerf less than 24 hours when I start to worry he’s controlling my mind.
To be fair, he’s been messing with my thoughts for a while before we finally sit down at Soho House in early December for drinks. Over the previous weeks, I’ve listened to him on podcasts and interviews as I’ve gone about my life. As I walk, headphoned, across the Northwestern campus where we both teach, I hear him talk about the possibility of directly connecting brains to machines. Cruising down the Edens, I hear him talk about influencing dreams through smells. Riding the moving walkway to Concourse C at O’Hare, I hear him wonder why April Fools’ Day is the one time when we treat the world with the skepticism it’s due. The 42-year-old neuroscientist has a 100-mile-an-hour French-Israeli accent, a gift for storytelling, and the ability to make me stop in the middle of the sidewalk and wonder if anything is real.
One of the first things Moran Cerf tells me, just minutes after we’ve greeted each other, is that when straight men and women shake hands upon meeting, they have a subconscious instinct to sniff their own hand afterward. (Keep in mind this was back in December, another human epoch.) This is regardless of whether the people are interested in each other, or available, or whether one is 19 and the other 85. A gay man would sniff his hand after meeting another man, and so on. I’m smiling and nodding, asking follow-up questions, but I’m also racking my brain, trying to remember what I did after we shook hands. It’s not so much that I’d be embarrassed if I did this thing — the fact that I’m a straight woman is not a well-kept secret — but rather that whatever I did, when I thought we were simply greeting each other, Cerf has absolutely noticed it. And in the past 10 minutes, what else have I done to betray skepticism, nervousness, insecurity, vanity, distraction?
Before Cerf went into science, he was a hacker. (“Hackers are anarchists,” he tells me — an observation that becomes more relevant the more I learn about him.) Essentially, his job now is to hack the human brain. I relate, in a small way. Part of being a fiction writer is taking note of other people’s body language and appearance and behavior — existing in the world simultaneously as participant and as vampire. But where I observe people for loose inspiration, Cerf is gathering data. And while I appreciate a good psychological hack, it’s unnerving to realize how easily you yourself might be read.
As our second glasses of wine arrive, Cerf asserts that his success (and he’s had a lot) has more to do with his skill as a communicator than any groundbreaking work as a scientist. He is, no surprise, charismatic. He’s also pointedly humble, and intent on asking me more questions than I ask him. It’s not long before I’m wondering how intentional that charisma is. It’s not fake — you can’t fake being interesting — but when you’re someone who’s keenly aware of the ways humans make first impressions, the ways we react to each other, and when the person sitting across from you is here to write an article on you, do you modify your behavior accordingly? I imagine Cerf would answer that we all would — that where I assume a clean division between sincerity and artifice, he sees a complicated network of instinctual and automatic decisions beyond our control.
During that first meeting with Cerf, I was mostly concerned that he was reading my mind. It’s the next day that I start to wonder if he’s manipulating it. It’s lunchtime, and he’s presenting to an ad hoc group of students in Northwestern’s Kellogg School of Management — some of whom look like they just stepped out of corporate training videos, and a few who look more like the hypercaffeinated writing students I know from the other side of campus. Several Kellogg professors are here as well, ready to learn about the brain.
The fact that Kellogg, of all places, is where Cerf has his lab is part of the story. Cerf’s primary neuropsychological research is in human single-neuron recording — that is, placing electrical nodes directly on the surface of the brain to record the activity of individual neurons. Yep, this involves opening and closing the skull. I ask with horror who’s having this done to them; the answer is people with epilepsy and neurological conditions who are already undergoing brain surgery (Cerf’s lab collaborates with a neurosurgeon in New York City) and are invested in the cure. With the electrodes already in place, they’re often willing to take part in extra studies. But as strides are made in the treatment of epilepsy, there might be even fewer opportunities to access actual living human brains.
Cerf discovered early on that there’s little interest in funding studies in which you ask the subject, say, whether she prefers the ice cream or the spinach and then watch what happens in her brain, which, yes, does indeed prefer the ice cream. What’s much more interesting is what a subject’s brain can tell you about attention, about how the mind responds to stimuli from, say, television ads. Of course, it’s all but impossible for a subject to tell you that herself. The second I ask if you’re paying attention to the movie you’re watching, I’ve broken the spell. If I wait till the movie’s over, you might have forgotten that although you loved the ending, you were distracted and confused in minute 20. Cerf realized that monitoring engagement with content in real time holds applications for educators, behavioral researchers (how engaged is someone on a date? or watching a political speech?) — and also for Hollywood, for online content providers, and for any number of other businesses with major money to make and to spend. This keeps Kellogg, and the businesses that hire its graduates, happy.
Cerf starts the talk with a magic trick. He shuffles a deck of cards and has someone in the front row call out the color they think the top card is. Black, the first student says. The next student calls a red card. Cerf sorts the cards, without showing them, into piles according to our guesses. He checks each card, occasionally asks us if we’re sure. The ones that audience members deemed black go on the left, the red on the right. Halfway through, he switches things up, putting the black cards on the right, the red on the left. Finally, he flips the cards over. We have been a hundred percent accurate.
I know we’re not psychic. The class continues, and I’m perplexed that the business students, as interested as they seem in the lecture Cerf has segued into about the illusion of choice, aren’t demanding he go back and explain the trick. It appears that he has no plan to let us in on his secrets; he just wants us to think about whether we were truly choosing freely. The obvious implication is that we weren’t, but what isn’t clear to me yet is whether he manipulated each of our choices (by moving his lips as if mouthing a B or an R? by hesitating just the right amount?) or whether our choices were simply circumscribed by sleight of hand.
He talks about some of his recent research. There’s a classic experiment in which a subject is shown two cards, each bearing the image of a human face, and is asked which is more appealing. A few moments later, the subject is shown the face they chose and asked to explain their reasoning. The trick is that a certain percentage of the time, the wrong face is shown, and yet the subject still comes up with justifications for their “choice.” The experiment has been around for decades, but Cerf’s lab is doing new data collection, monitoring neural activity during the moments when people explain both their actual and supposed choices.
Knowing what happens in the brain in these moments, Cerf tells the business students, has implications for product marketing, for elections, for Matrix-like manipulations of reality. “We can actually go into your memory,” he says, “and change things, and you will create a new narrative for yourself. And you will never question this understanding of who you are. You tell yourself a story that this was always what you wanted.” He doesn’t mean that he can trick you into thinking you’ve always been a Republican, he adds, but that people can be moved slightly on issues like climate change or gun control by being told they responded to a survey in a way they didn’t and then being asked to justify their choices. “We give so much skepticism now to the news,” he says, “and yet so little skepticism to anything that originates with us.” We trust our own memories implicitly. But because memories are often faulty and sometimes prone to manipulation, perhaps we should be approaching our own minds as if it’s always April Fools’ Day.
I’ve taught graduate creative writing at Northwestern for five years, and coming to Kellogg has shown me a whole new side of campus. I’m used to Hogwarts-looking buildings where students still lug backpacks full of books to class, but the classrooms here are glass-walled and sparkling. There’s an impressive receptionist-to-coffee-station-to-smartboard ratio. Copious Potbelly has been carted in for lunch. It’s safe to guess that each of these business students will earn more in the next five years than all of my writing students combined. They’ll do so in part by making us want things we don’t have, convincing markets and populations of the narratives that serve their own businesses.
After class, I make Cerf promise to let me in on the magic trick the next time we talk. I notice I’m the only one asking. Perhaps the business students are content not to know, or perhaps this is old hat to them, the idea that we’re here to manipulate and be manipulated. And what I’m wondering, but not asking, is whether any of the choices I’m making right now — the way I’m standing, how I’m wearing my scarf, the words I’m using, what I’m doing with my hands — are truly within my control. I mean this both metaphysically (what is choice? do we have free will?) and literally (am I mirroring the way Cerf is standing? is it because he’s making me?).
As a writer, I’m no stranger to manipulation. Even now, the image in your mind of Moran Cerf is the one I’ve planted there, the one I’ve curated for you. And since this is a written piece, you’re perhaps approaching it with your antennae up. You may question my perceptions and judgments, wonder if I’m biased toward a fellow academic, too generous toward someone I find personally compelling. But you likely would not have those same reservations about your own in-person perceptions of Cerf, and he would argue that you should. A T-shirt his students made for him reads “Don’t Believe Everything You Think.”
By any measure, Moran Cerf has led an extraordinary life. Born in France and raised in Israel, he entered a select school for the performing arts at age 6 as a ballet and theater prodigy. From 7 to 14, he was one of 20 of the school’s students featured on a weekly educational television show (the kids would sing, cook, and act in segments related to the theme of the week), making him a nationally recognized face. At 18, he left the school a year early to study math and science (“I was the black sheep of the group,” he says of his small cohort, almost all of whom now make a living in the arts) and then began his obligatory service in the Israeli army. Already adept at rudimentary hacking (as a kid, he found a trick for adding extra lives in Super Mario and later used dial-up internet to learn from early hackers all over the world), he entered intelligence training and spent four and a half years learning cybersecurity in an intel unit full of fellow hackers.
After the army came a university degree in physics, work for a tech company, and a master’s in philosophy. He started his own company in Israel as a white-hat hacker in 2001, working on assignment for companies like Amazon, Facebook, Google, YouTube, and Bank of America. They’d hire him to breach their own security to test the strength of their firewalls. “I have a lot of bank robberies on my sleeve,” he says.
You know: Just your average ballet-dancing child TV star bank robber philosopher hacker.
His business was eventually bought out by the Silicon Valley cybersecurity company Imperva (Cerf naively chose a higher salary over more equity; he lives quite well still off his stocks, but life would be different if he’d opted for the opposite), and he spent the next few years traveling the world on behalf of the company, often attending lectures during his free evenings. On one such night in California in early 2004, he met Francis Crick, the man who cracked the genetic code in the 1950s. In the last years of his life, Crick had been studying human consciousness. The brain was the next frontier, he told Cerf, the next great black box. Cerf was wasting his time working for the corporate world, Crick said. A protégé of Crick’s was placing electrodes on human brains; if Cerf was interested, Crick would write him a letter of recommendation. And so, at 28, Cerf entered the doctoral program in neuroscience at Caltech.
The group research he chose to do there — single-neuron recording — was both high risk and high yield. A major paper on the group’s findings came out just as Cerf graduated, and he was quick to land a faculty post at New York University and eventually at Northwestern. Although his research has taken him in additional directions, the electrode-based brain mapping is still what he’s most frequently asked to lecture on.
Cerf is, perhaps, a case study in the argument for giving scientists an arts education. The friendships he’s maintained with actors from his student days (many live in Hollywood now) have fostered an interest in storytelling. He has served as a consultant on several TV shows (Bull, Mr. Robot, and the short-lived Limitless and Falling Water), and he teaches workshops on science in film at the American Film Institute. His 2013 TED Talk has reached tens of thousands more people than any scientific paper ever would. He has won the Moth storytelling competition multiple times. This is a man who understands how to put on a show. “TED Talks require you to deliver one idea,” he says. “You need a sound bite. You have to dumb it down, make it funny. There’s a price, and the price is the quality. People give maybe one hour a year to science.”
I eventually ask him outright if what he knows about our perceptions tempts him to bend reality to his benefit. “You mean if it makes me better at negotiation or dates?” he says. “The answer is, unfortunately, no. All these things work in theory, but it’s very hard to implement them in the moment.”
Weeks after the class at Kellogg, we meet in Cerf’s West Loop loft. It’s the apartment of a grown professor, but it’s also clearly the apartment of the hacker teenager who still bubbles to the surface when Cerf is excited. He’s unmarried, no kids, so the place is entirely his — like walking into the rumpus room of his brain. There’s an Einstein poster, and a collection of broken clocks, and a telescope, and a giant photo of Phineas Gage, the 19th-century railroad worker whose asymmetrical brain injury made him a foundational neurological case study. The Pac-Man art on the wall is something Cerf made himself. He has a sword.
Every single thing that comes out of Cerf’s mouth is both fascinating and disconcerting. He tells me about an Israeli woman he knows who developed a painful sensitivity to electronic radiation and could sense where in the room it was coming from. Fair enough. Then he says, “What if we converted that pain to some other sensation? To joy?” This is related to a project he’s working on now, he says. “We sense things, the brain receives them in a certain way. But what if we could shift that reception? We could connect a smell to math. And then if I ask you to do 111 times 48, instead of doing the logical math, you could say, OK, 111 smells to me like a lemon, and 48 smells to me like papaya, and lemon and papaya together smell to me like this, and this is clearly 732.” I’m wildly and inexplicably relieved, later, to confirm that 111 times 48 is miles away from 732.
He also mentions the applications of artificial intelligence for writing. The best-selling author in the world right now, he says, is a computer program created by a professor of marketing named Philip M. Parker. It creates nonfiction books on demand by culling the internet for information on whatever topic the buyer has searched for, then publishes them under Parker’s name. (I couldn’t confirm the bestseller claim, but a quick search turned up more than 50,000 titles by Professor Parker.) Most people who order these books have no idea that what they’re getting has been generated by a computer. As a writer, I am rankled by this, and not just because of the plagiarism implications. I don’t love the suggestion that human creativity is something replicable, or even available for analysis. Body language, shopping behavior, even dreams — map those all you want. But not the process of creation.
What I really need right now is for him to show me the magic trick again. I’ve avoided Googling it, perhaps because that would be cheating, and perhaps because I’m afraid the trick will turn out to be mundane. Cerf obliges, producing a deck of cards and asking me again to guess if the cards on top are red or black, and again (I must really be psychic!) my guesses are a hundred percent correct. By way of explanation, he tells me that the secret to figuring out any magic trick is to look for the steps that are not necessary: “This is how hackers work, how scientists work, and how magicians work.” We go through the trick again, and I see it. There’s absolutely no reason, for instance, that he should have to switch, halfway through, from red on the right to red on the left. Something’s up, and it’s not about mind control; it’s about sleight of hand. I won’t divulge more of the secret than that, but I’ll say that no, Cerf was not (in this one instance, at least) controlling my brain. The idea that I had free will was an illusion, but of the more classic, prestidigitational variety.
Cerf trains his students to look for these kinds of Easter eggs too. In each of his regular class meetings, he drops a password — it may be in conversation or a case example, or written in the corner of the board, or the wrong mathematical answer he intentionally lands on, waiting to be corrected — and any student who notices it and types it into an online portal gets the answer sheet for the day’s assignments, saving them hours of work. Only once has a student ever guessed the password, he says. “But they try. They learn to try better. I keep a record of all the passwords they’ve tried.” The point is not that they hang on his every word, but that they train their minds to seek out the odd. “It’s inspired by Sherlock Holmes,” he says. “It’s a philosophy that everything matters. Even the mistakes.” It does occur to me to wonder if the number 732 is somehow the key to the universe.
In early February, I again cross campus to Kellogg, this time to be a guinea pig. I told Cerf I wanted to know what it’s like to have your brain scanned (from the outside — I like my skull the way it is). We meet in his office, and he fits me with something like a swim cap, dotted with 34 electrodes that transmit my brain waves to a monitor, where they look a lot like an EKG readout. The electrodes divide my brain into a hundred million clusters of activity, he explains, and while I don’t fully understand the biomechanics of this, I trust him. At least he’s not sawing me open.
He’s running that face experiment on me, the one where he asks which person I prefer. This time it’s on a laptop, with the computer occasionally switching my choices. Since I’m onto the trick, I catch some, but probably not all, of the swap-outs. I try my best to be honest about why I’ve preferred which faces; I try not to overthink it. When we’re done, I make mention of his not being able to use my data because I knew how the experiment worked. No, he explains, they’re actually collecting lots of data points I knew nothing about: whether subjects tend to prefer the photo on the left or the right, for instance, and how long it takes to decide. Then he looks at me quizzically. “You know, you gave some really strange answers,” he says. “Most people just talk about how they like someone’s eyes.”
Later, listening back to my recording of the session, I realize that in my attempts at visceral honesty, I might have betrayed some odd things about myself. What came out of my mouth at one point was “I like her because she looks like a babysitter who got murdered and her body wound up in the pool.”
It’s gratifying, at least a little bit: The stranger my brain is, the less likely it is that my creative process can be replicated by a robot. This is what I tell myself, anyway.
Cerf is the first to agree that just because something can be done (the manipulating of dreams, the planting of false memories, the AI’ing of authors), doesn’t mean it should be. “It’s a conflict I have,” he says, “and all scientists should have. We’re in a tough position. We’re the ones opening Pandora’s box. And we tell people, ‘It’s Pandora’s box. You decide if you want it open or not, but we’re going to create the key.’ I don’t think scientists should be the deciders.”
In our conversations, Cerf seems genuinely torn between the research itself and the ways it might be used. I question his sincerity — he’s taught me to do that! — but I do believe he’s conflicted. “Deepfakes and fake news are hints to the fact that we can’t trust things the way we did before,” he says. “It’s going to get harder and harder. I’m surprised that no one talks about it. I would expect that someone would ask the Democratic candidates their view on AI, their view on genetic editing. And no one asks it. The only question on technology in the last election was Hillary’s emails.”
This is part of the reason he’s so keen on giving talks, on letting the public know what’s going on in the labs. He tries to present the dilemmas to the audience, but he’s careful not to be political or prescriptive in his take. “It’s easy to be an academic; you can stay pure and people will let you talk about whatever you want. But if you have a public point of view, then people will label you.” But sustaining a neutral, academic stance is dangerous too. “I’ve convinced Kellogg and my colleagues and the world that I’m safe, that because I’m doing it, it’s good,” he says. But in fact he believes the opposite is true: “Because I’m the one doing it, I can tell you with authority that it’s not good.”
And so the guy who’s spent the past decade researching attention and influence is still trying to figure out the best way to use his own influence without losing our attention.
A few days after I last met with Cerf this winter, my kids and I ate at a family-friendly restaurant with a roving resident magician. He came to our table and did some pretty fantastic, if standard, sleight of hand (the card you chose has a giant red dot on it now, the card you chose is rising from the deck, unbidden). I was watching, as Cerf had told me to, for the small things that made no sense — the unnecessary gesture, the extraneous move — and there they were. Had this man really needed to point out his own necktie? There was the trick.
If scientists look for the slight catch, the anomaly, the weird data blip because that’s where discovery is most likely to be, then it might be true as well that this is the tell for more personal kinds of discovery. It’s what I look for as a novelist too: the moment in someone’s life when things don’t quite sit right, when something’s got to give. That’s the best place to start a story because that’s when someone’s about to change. It makes me wonder if I’m meeting Moran Cerf right before another major pivot in his life. If the next time we hear about him, he’ll be doing something so wildly different that his having been a Northwestern professor will seem as improbable as his having once been a ballet dancer.
In April, I emailed Cerf to see how he was doing in quarantine and to check some things about his work. It was late, but I had a feeling he’d be up. (Imagining Moran Cerf sleeping is like imagining a whirlpool at rest.)
He wrote back immediately and said he was working on a project that could use a “fingerprint” of your brain as the ultimate password: “a password that you can use but not know.” Using what’s in Pandora’s box, in other words, without quite opening it.