Proving precognition, programming a screenwriter, and other tales from the field

Scientists share their favorite stories.

Memory athlete to memory expert

memory experts
"Anyone can learn to improve their memory, but it’s not like they’ll remember everything."Peter Oumanski

↑ Boris Nikolai Konrad, guest researcher at the Donders Centre for Cognitive Neuroimaging, Radboud University, Netherlands I started training my memory to help with college exams. A few years later, in 2003, I became a memory athlete. During one type of event, you stare at a list of digits for an hour, then over the next two hours, you write down every one you can recall, in order. Most pros turn digits into a mental image—a 3 might become a billboard—then stick it in a known location, a method known as a Memory Palace. You might start out with only a few familiar sites, but you can gather more. I have about 80 palaces with 50 locations in each; when I go someplace new, I often explore to make a fresh palace. When I finished my master’s ­degree, I decided to research the effects of these exercises, ­comparing the brains of memory athletes and noncompetitors. There’s nothing special about an athlete’s brain. Even scans of my own showed me to be exactly ­average, which was a little disappointing. We did find a difference in brain activity: When competitors are trying to remember more effectively, they employ different areas of the brain—such as those for visualization—in unison. Anyone can learn to improve their memory, but it’s not like they’ll remember everything. For example, we showed test subjects a set of words, and told them to remember some but forget others. Later, the memory athletes could recall the targeted words better than the nonpros. When we asked for the words we had told them to forget, they got angry. They said: ”It’s not fair! If you’d told me to remember them, I could have!” When we finally got them to stop yelling, they did as poorly on those as everyone else. As told to Rachel Feltman

Crashing crow funerals

crows
"..they might fall silent and stare at the dead bird, almost like they’re in mourning. Watching this behavior is—“moving” is not the right word. It’s startling and incredible.Peter Oumanski

↑ Kaeli Swift, graduate student at the University of Washington When one crow sees another dead on the ground, it caws an alarm. Then others—five to six on average, but in rare cases as many as 60—fly in and perch on branches, cawing. In anecdotal reports, they sometimes place objects like sticks over the body as if burying it. Then they might fall silent and stare at the dead bird, almost like they’re in mourning. Watching this behavior is—“moving” is not the right word. It’s startling and incredible. I started studying this activity to understand if crows just do this as a response to danger, or if their behavior serves some other function. To do that, I stage my own crow funerals in downtown Seattle. Crows are smart: They can recognize ­human faces. To deal with this, my researchers and I wore identical masks so the birds would think we were all the same person. Then we laid a stuffed crow on the ground and watched the birds perform their ritual. We found that the funerals probably serve to identify potential new predators—like us—and warn the group about them. Watching crow funerals is not like watching humans mourn. I know what a human feels; I can’t empathize with crows. My job as a scientist is to not make assumptions about their feelings. As long as we live in cities, crows will be part of our lives. I hope my research helps people realize that these ubiquitous and annoying animals are actually kind of cool. As told to Ellen Airhart

I loved writing the SAT

SAT writing
"Actual humans have to write these questions, but most people can’t believe it."Peter Oumanski

↑ Jim Patterson, Director of Assessment Design and Development at the College Board When I was an English teacher, I got familiar with standardized tests, and in the mid-’90s, I joined the industry and started writing questions. Today, I supervise the people who supervise the writers. But I still love picking apart a good reading passage, whether it’s from classic American literature, historical documents, or contemporary novels. To find a reading-comprehension excerpt, you look for a nugget that’s easy enough for high schoolers to understand but complex enough that you can mine it for a variety of questions and difficulty levels. For a passage about, say, a balloonist, you might ask how the narrator’s attitude shifts, or how in line 50, “bearing” most nearly means ______. Actual humans have to write these questions, but most people can’t believe it. Their initial reaction is surprise. Years ago, at my 10th high school reunion, I told someone I hadn’t seen since graduation what I did for a living. She made a face, paused, then said, “I can definitely see you doing that.” I took that as a compliment. As told to Sara Chodosh

How to program a screenwriter

screenwriter
"Sharp enlisted technologist Ross Goodwin to build what might be the first script-writing artificial intelligence."Peter Oumanski

↑ Oscar Sharp, Director; Ross Goodwin, Technologist When director Oscar Sharp entered Sci-Fi-London’s 48 Hours Film Challenge in 2016, he had to write, shoot, and edit a movie in two days. To up his creative game, Sharp enlisted technologist Ross Goodwin to build what might be the first script-writing artificial intelligence. This recurrent neural network, which eventually named itself Benjamin, spat out the four-page screenplay that became the seven-minute Sunspring. Then things got weird. 1. Learning To train a neural network that writes text one letter at a time, Goodwin fed it nearly 1,000 sci-fi scripts, including The X-Files and Blade Runner. 2. Writing Benjamin churned out a jabberwocky of a script. Although it contained no character information, the actors interpreted the tale as a love triangle. 3. Staging Sharp needed to make sense of some very odd actions: For a character “taking his eyes from his mouth,” the director had the actor spit out a prop eyeball. 4. Screening Audiences loved that an AI had written the script—though some considered it mere word salad. Out of more than 100 competitors, Sunspring made the top 10. 5. Walking and talking For Sharp’s next project, Goodwin trained Benjamin to write like screenwriter Aaron Sorkin. Its latest line: “The president wants to do it for me.” As told to Jason Lederman

Proving precognition

precognition
"Suddenly, I was known for paranormal research."Peter Oumanski

↑ Daryl Bem, Professor Emeritus at Cornell University As a social psychologist for more than 50 years, I’ve studied topics ­including sexual orientation and personality theory. But I became known for my work on precognition, anticipating the future. I got hooked on this stuff as a kid when I saw a “mind reader” named Joseph Dunninger on TV. In high school, I started performing my own mentalism shows. When I became a professor, I showed my students the same tricks to demonstrate why witnesses make mistakes. In 1985, the Parapsychological Association, researchers who study extrasensory perception, asked me to perform at a conference. On stage, I correctly listed the contents of a packed box without opening it. It was pure swindle, but the point was not to demonstrate that I could read minds; it was to show how easily fraud could ruin their tests. One of the scientists asked me to make his lab’s telepathy experiments cheat-proof. He got real results—unexplained communication between subjects in separate rooms—so I helped him get published in an academic journal. But he died nine days before our write-up was accepted. Suddenly, I was known for paranormal research. Starting in 2000, I’d found existence of precognition in nine experiments, by reversing the chronology of psychological tests. In one classic study, people take longer to recognize an image as pleasant when the word “ugly” flashes by before it. I flipped the order of events, so subjects viewed the image before seeing the subliminal word. I still found the same effect, showing they could “feel” the future. Eight of my experiments are replicable, and I’m redesigning the ninth one so everyone can get results from it too. In a way, I’m betting my career on it. As told to Amy Schellenbaum

Looking for life in all the wrong places

Alien life
No phone homePeter Oumanski

↑ Jill Tarter, Astronomer and Bernard M. Oliver Chair at the SETI Institute Since the 1970s, I’ve been pointing optical and radio telescopes at the sky to look for engineered signals from other planets—signs of extraterrestrial intelligence. In 1984, I broadened that search when I helped found the SETI Institute. At SETI, we built tools to look for signals we don’t think nature can produce. One time, in 1997, at West Virginia’s Green Bank Observatory, my team noticed a strange signal from a distant star. On a graph of frequencies over time, it looked like a picket fence: spikes with equal spacing between each one. It was such an adrenaline rush. When you think you’ve found an alien transmission, it’s heart-stopping. I wrote a computer program to take another look at the data—but I didn’t take the time to format it properly. Because of that, I misinterpreted the results. Hours later, we realized our telescope was picking up transmissions from the European Solar and Heliospheric Observatory passing in front of the sun. If I had been more level-headed and read the output correctly, I could have saved all that time and frustration. A BBC crew was filming us that day, and I think they were more crushed than we were. To make it worse, we had alerted some colleagues in California to watch that same star. We had gone off to bed after we sussed out it wasn’t a signal. But we forgot to tell them! They called back at 2 a.m., and they didn’t think it was funny at all when we told them it was a false positive. I had a number of fences to mend. As told to Mary Beth Griggs

Chamber, made

maze
Mind gamesPeter Oumanski

↑ Andrew Parr, owner of New Escape Room Designs Inc. I’m a full-time music teacher, but on the side, I design puzzles for escape rooms. You know, the physical adventures where small groups find clues and solve riddles to get out of a locked chamber. You work your way to a final key that unlocks the door, allowing you to escape. As a creator, I have to assemble a room that is fun yet hard—and most important, fair. The challenge is to imagine what a group of strangers might do to follow my puzzle. That’s tricky, because when I’m designing a game, I’m starting with the answer and working backward—for players, it’s reversed. So I have to make sure each step makes logical sense. I’ve gone undercover into escape rooms to play my own games and watch what people do. I’ve found I don’t need to write incredibly challenging riddles. A player starting at the beginning of a mysterious set of puzzles will automatically perceive that mystery as hard, even if at the end they look back and say: “This was completely solvable. The clues were all there; I just didn’t see how they combined.” I also have failed to figure out a set of ­solvable riddles, when some friends and I played a room I didn’t write. We had found four animal statues, and the hints told us to place them on pedestals. Turns out, they also had to be facing a certain direction because they had magnets inside. When lined up correctly, they would trigger a box to open. The clues were as precise as they could be. They gave hints like “the monkey is not facing the same way as the lizard.” But under pressure to finish on time, we got sloppy. We couldn’t decipher what the tips meant. It wasn’t the game’s fault—and that drove us nuts. But that’s how the room works. As long as the pieces are there, I don’t mind being defeated by a good puzzle. As told to Claire Maldarelli

I spy with my little eye

satellites
"in one year, our informal group recorded 18,000 observations of 200 spy satellites."Peter Oumanski

↑ Ted Molczan, energy-conservation consultant and secret-satellite tracker In summer 1968, I was standing in my driveway in Hamilton, ­Ontario, when I spotted a brilliant light in the sky and knew it was a satellite. I was a pathetic math student—I had just failed my first year of high school. But I was keenly interested in science and space. So I tried to calculate “my” satellite’s orbit. By sheer luck, my shaky math worked out: The next night, the spacecraft reappeared right where I had predicted. It turned out to be Echo II, one of NASA’s early communication satellites. Identifying it had a profound effect on me. I’ve since spent a good part of my life as an amateur satellite spotter. Plenty of hobbyists follow the thousands of unclassified commercial, military, and scientific orbiters. But I spot spy satellites. About 20 of us around the world track roughly 400 of the crafts. Finding unclassified satellites is pretty easy—governments release data about their orbits. But my group can still track the ­objects that operate in secrecy because we know about their launches: Official agencies announce them (they’re hard to hide!) and issue warnings that tell us where parts of a launch vehicle will fall to Earth. With that, we use computer programs or even hand calculators to determine the classified object’s orbital plane. Then we can view it with binoculars or high-quality cameras, and time it with stopwatches. In one year, our informal group recorded 18,000 observations of 200 spy satellites. The work we do can help journalists to report on government activities and political scientists to shape policy. They rely on our findings because most of the people who launch classified satellites don’t leak information. Of course, I wouldn’t share any findings that might threaten the national security of ­Canada or its allies. I’ve seen and figured out a few things that I might never make public. As told to Mary Beth Griggs

How you bank a brain

brain bank
"When a donor dies, we have only 48 hours to process their gray matter."Peter Oumanski

↑ Samantha Rice, Assistant Director of Histology, Harvard Brain Tissue Resource Center For about a year, I’ve worked at the Harvard Brain Tissue Resource Center, which stores around 3,000 brains for research. When a donor dies, we have only 48 hours to process their gray matter. An on-call pathologist—we have an extensive U.S. network—removes the organ, packs it in an ice-filled plastic foam box, and ships it to us. Then a courier picks it up at a local airport and rushes it to our labs, where a dissectionist like me steps in. After I photograph the donation, weigh it, and examine it, I cut it in half. Brains are squishy like jelly, so to keep the cells stable, I put one hemisphere into a chemical solution that firms it up. I dissect the other side into hundreds of little pieces and freeze them in liquid nitrogen. We send these samples all over the world; each organ provides enough tissue for hundreds of potential studies. Every brain I handle is unique, and helps scientists understand disorders and diseases such as schizophrenia and Parkinson’s. I plan to donate my brain, along with the rest of my body, and I encourage others to as well. As told to Nicole Wetsman

It's surprisingly hard to trick an infant

baby scientists
"...babies have sophisticated knowledge about the world from birth."Peter Oumanski

↑ Aimee Stahl, Developmental Psychologist and Baby Magician at The College of New Jersey Many people assume that ­because babies can’t walk and talk, they don’t know much. But I study infant cognition, and ­babies have sophisticated knowledge about the world from birth. Working with infants is ­extremely fun but also challenging: You can’t ask what they’re thinking. We know babies look longer at things they find surprising. This tells us whether objects that seem to break physical laws, like a ball that passes through a wall, defies expectations. To pull off these tricks, we set up a puppet stage with a ramp leading to a wall and a screen that hides part of the view. We release a ball down the ramp, it rolls behind the screen, and we secretly place it on the other side of the wall. When we lift the screen, the baby thinks the ball passed through the wall. Afterward, they often try to bang the ball against a solid object, as if it should pass through that too. As told to Claire Maldarelli

This article was originally published in the Spring 2018 Intelligence issue of Popular Science.