2012 Thesis Excerpts
The Twitching Eye: REM Sleep and the Emotional Brain
One night in December 1951, an eight-year-old boy fell asleep in his father’s lab. Chemicals spurted across a synapse in the boy’s brainstem that night, as they typically do in the human brain about ninety minutes after falling asleep. In response to this tidal shift in brain-juices, the boy’s eyes began to twitch; his mind awoke while he slept. He dreamed.
What was new that night was that someone was watching sleeping eyes and brain waves at the same time. The boy’s father, Eugene Aserinsky, a 30-year-old physiology graduate student at the University of Chicago, was recording eye-movements and electrical activity from his son’s sleeping brain in search of a doctoral project.
Eugene’s pregnant wife, Sylvia, cared for their son, Armond, in a chilly dormitory, renovated from a World War II barracks. Sylvia battled manic depression while her husband toiled in the lab at night. They were poor, and Eugene Aserinsky was a name no one knew. Aserinsky’s boss, Nathaniel Kleitman, the world’s preeminent authority on sleep at the time, believed that sleep was the brain’s off state – a period of rest to recuperate the body. “Sleep is to waking as ice is to water,” Kleitman would say. But on this night in the basement of Abbot Hall, our view of the sleeping brain changed.
When Eugene Aserinsky saw Armond’s brainwaves spiking in rapid jerks on the EEG, he thought the boy must have woken up. Ink-pens wiggled in time with the boy’s brainwaves, making a soft scribbling sound, the output of the machine recording electrical activity from electrodes on his scalp. Wavy lines appeared on a spool of white paper, as his neurons fired in synch. Slow-moving waves meant deep sleep. When the pens scribbled fast, it usually indicated the boy was awake. But then Eugene noticed Armond’s eyes: lids closed, rapidly moving back and forth. The brain looked awake, but not the boy.
Since sleep scientists like Kleitman only expected to see two stages—“Awake” and “Asleep,” like the brain’s on and off modes—they had never recorded far past sleep onset, when the brain slowed down. They assumed the rest of the night would look the same: the synchronous brain peacefully fluctuating like an ocean, slow waves rhythmically lapping up on the shore of the scalp from the depths of sleep. But that night in Chicago, when Eugene Aserinsky watched his sleeping son’s brainwaves over a longer period, he glimpsed a new territory no one had seen before: the time at night when the brain awakes while we sleep, the eyes move and we dream. In the world where Armond Aserinsky woke up, his father had discovered what he called rapid eye movement (REM) sleep, and a new field of science was born.
Full thesis here: http://dspace.mit.edu/handle/1721.1/77473
The idea that artificial intelligence systems are constrained by what their programmers put into them is at the heart of what researchers call “narrow AI.” But there is an alternative: “strong AI,” a term that John Searle coined despite the fact that he doesn’t believe it’s possible. Theoretically, strong AI would think like a human thinks, adapt to new problems, learn new skills, improve itself, and be flexible in ways traditional computers are completely incapable—as Searle puts it, a computer that has a mind “in exactly the same sense that human beings have minds.” For all intents and purposes strong AI is a mind, only built from silicon and computer code instead of neurons and synapses.
It’s hard not to hyperbolize when imagining the applications or abilities of such a computer. It would combine the efficiency and speed of traditional computers with the flexibility and adaptability of the human brain. In theory, given mechanisms for perceiving and interacting with the world, such a machine would be able to do anything a human could do, from tying a shoe to doing original scientific research.
The most ambitious and imaginative AI researchers believe that strong AI would be capable of solving some of the most difficult problems humans face, problems that arise from staggeringly complicated sets of interconnected variables unfathomable by the human brain. Such a machine could end poverty, or hunger, or disease. Such a machine could end war.
But back to earth; the reality isn’t nearly so impressive. Instead of grappling with such fundamental and intractable questions as what is the mind, most AI researchers take the comparatively simple approach of solving specific tasks. How do we find things on the internet? How can we get a driver from Cincinnati to Des Moines? How can we win at Jeopardy!? And how can we make computers do these things faster and more efficiently than a human can? All existing artificial intelligence is narrow—all of it, even the most sophisticated and intelligent-seeming programs out there.
But there is a small and dedicated community of researchers who seek to advance the field in fundamentally new ways, scientists who are less interested in the narrow applications of AI than they are in understanding the mind, and strengthening AI based on that model. They concentrate on foundational questions and make substantive contributions to AI as a science while simultaneously expanding our understanding of what makes humans human.
Their approaches vary widely. Some think that biology is the way forward, that a neuron-by-neuron recreation of the brain is the surest way to produce intelligence. Others ask questions about what makes humans different from other intelligent species, and try to engineer systems that reproduce those differences. Some scientists think that describing the mind mathematically will provide a precise blueprint for intelligence. And yet others take cues from a variety of places, the pragmatism of business and the hard math of cognitive psychology and the rigid determinism of computer science.
But all of their research shares a common ancestor: the human mind. They believe it is the best model we have for flexibility, for adaptability, for generalized intelligence capable of surviving and growing in a complex world. We are the result of millions of years of evolutionary progress and the best (and only!) example we have of a system that gives rise to real intelligence, even if we can’t say exactly what that means.
Their quest to build a smarter computer is, at its core, the quest to understand the human mind….
Full thesis here: http://dspace.mit.edu/handle/1721.1/76137
As he watched Florida recede in the distance, thoughts of the journey behind and the one lying before Chris German filled a hopeful mind. “Deep joy,” describes German, reflecting on his outlook at the start of the expedition.
This voyage to the Mid-Cayman Rise (MCR) was planned to last 22 days. In 28 years of exploring since he began his PhD studies at the University of Cambridge, German has spent around a thousand days at sea, about on par for a modern day oceanographer. It’s a life of discovery and adventure but it also demands sacrifices—while at sea he’s missed his wife’s birthday (twice), his son Martin’s 18th birthday, and his daughter Helen’s 16th birthday (while in the Antarctic).
Now on Atlantis, bearing down on potentially the most significant sampling of hydrothermal vent sites on the planet, German hoped real discoveries would be waiting for the team somewhere about 400 miles south of Havana.
In addition to grabbing temperature readings, biological samples and close-up shots of the vents, German wanted to study the MCR for what it can tell us about Earth’s history. Since the MCR is an ultraslow spreading ridge, it acts differently than its faster counterparts. Usually rocks get older the farther you go from the spreading center of a mid-ocean ridge. That’s because new seafloor is being created by magma erupting and then cooling and molding together with the older surrounding seafloor. But at ultraslow spreading ridges, one side of a tectonic plate is older and heavier than another and ends up sinking—or subducting—under another plate. As it sinks, it pulls the other side of the plate out from underneath the next plate over—a bit like pulling one book out of a stack of books that have tilted over on their sides on a bookshelf. The side that gets pulled out is even older crust—old enough, potentially, to have been around during the brief but legendary lifetime of our good friend, Earth Life Form #1.
At 9:45 p.m. on January 8, Atlantis reached her destination above a hydrothermal vent site called Von Damm….On the deck beneath two 42,000-pound capacity cranes sat the Remotely Operated Vehicle (ROV) Jason, named after an ancient Greek mythological hero famous for leading a group of adventurers called the Argonauts. Jason is the most capable U.S. academic research vessel roaming the seas today, able to reach 99% of the world’s seafloor at depths down to 21,000 feet.
Tethered to Jason is it’s partner vehicle Medea, which serves as something of a shock absorber for the Greek hero. Any yanks on the line from Atlantis are transferred down to Medea, which maintains enough slack in the fiberoptic cable to allow Jason to roam around within 115 feet of Medea. In Greek mythology, Medea was Jason’s wife.
The next day, the engineering team had Jason ready to go by noon, but a few technical difficulties pushed the inaugural dive back a few hours. By 4:30 p.m., Jason was on the ocean bottom with all systems running. Its first photos from the deep were of weathered rock outcrops, a few fish and shrimp, mussels and tubeworms. Data were displayed in an operations van—a mobile boxcar of screens and instrument displays for monitoring Jason’s progress that shared the deck of Atlantis with a host of equipment. Using its manipulator arms, Jason collected five tubeworms, a clam, and a sea cucumber.
The science teams started working their round-the-clock shifts. Twenty-three people from nine institutions in three countries made up seven science teams. There was the Carbon Team made of Max Coleman and Sarah Bennett from NASA’s JPL. They wanted to track the state of carbon as it exits from the vent sites and how it gets altered through the food chain. “We’re looking at the whole carbon cycle, from abiotic carbon production deep within the crust, to biotic carbon production through chemosynthesis,” says Bennett. “If hydrothermal systems were to exist on Europa, it’s studies like this that may help us to calculate the biomass that may exist.”
Just about everything seems to revolve around carbon. It’s an organic element, so the biologists study it because it tells you something about how life has evolved there. It’s preserved in rocks, so the geologists want to measure it to figure out the ages of their samples. It gets spouted from superheated gases, so the chemists want to know what it mixes with.
After Jason’s first dive was declared a success, the crew of Atlantis brought the ROV back up to the surface. Securing Jason and Medea to the deck and distributing samples among eager science teams, Atlantis made for the second target, named the Piccard vent site (named after late undersea explorer Jacques Piccard), merely an hour’s drive away…..
Full thesis here: http://dspace.mit.edu/handle/1721.1/76138
Interview with an Octopus
“I dip a few fingers into the water. It is cold, but given this octopus came from just off the shore of Vancouver, B.C., I’m not sure why I had expected a tropical temperature….My fingertips find the very end of one of her arms, tapered into a fine point, her suckers here only small dots. She does not recognize my taste and recoils. But I am insistent, touching her a bit more boldly this time. Her skin is far from slimy—more like wet velvet than the sticky surface I imagined. She is a creature divided, paying rapt attention to Brorsen with half her arms while deciding what to do with me with the others. She clearly is not thrilled to have to deal with me, but tolerates my touch, more or less.
Then, a lucky move: Brorsen fumbles one of the shrimp he has been feeding her. I pick it up and put it in her suckers a little further towards her center. While she sometimes feeds herself by moving her arm to put the shrimp closest to her mouth where all eight arms meet, this time she chooses to shuffle the shrimp up her arm, passing it sucker to sucker. This little offering seems to have earned me some trust, and when I go to put my hand near her larger suckers this time, she greets me with enthusiasm, sucking on my fingers and winding her arm up around my wrist.
The reason I am here visiting Brorsen is because of something extraordinary that happened to him a few years ago.
As an aquarist at Monterey, Brorsen rotates exhibits once a year. During one rotation with the octopuses several years ago, Brorsen became particularly attached to one male octopus. They routinely played games around feeding time, including “catch the water”: a game of Brorsen’s invention which involved using his hand to squirt water in the air for the octopus to catch in its suckers. When Brorsen was rotated away to another exhibit, he grudgingly passed on the care of his octopuses to new aquarist.
But a few weeks later, Brorsen found he missed his octopuses and returned to the exhibit to say a quick hello. He dipped his hand into the familiar tank, and the young male reached up an arm to touch Brorsen’s hand. Octopuses are skilled at recognition, able to identify people by sight but even more easily by tasting with their suckers.
The moment the octopus’ sucker touched Brorsen’s hand, the octopus launched itself out of the tank. All eight arms flew towards unsuspecting Brorsen, wrapping around his neck and chest. Brorsen had to gently pry the enthusiastic, and sticky, octopus off his body, placing it back in its tank.
I have a difficult time imagining exactly how I would react to a 65 pound giant octopus flinging itself at my face. But as Brorsen recounts the story, he smiles. He knew the gesture was not at all aggressive. If anything, he says, “It was like a hug.” A wet, sticky, eight-legged hug.
But even as Brorsen tells the story with obvious delight, he holds back. He hesitates to characterize what happened as anything more than “like a hug.” To call it an actual hug would be projecting, he says, and he hesitates to make assumptions.”
Full thesis here: http://dspace.mit.edu/handle/1721.1/76139
Flashback: The Return of Psychedelic Medicine
In April of 1943, Swiss scientist Albert Hoffman synthesized a small sample of a compound called lysergic acid diethylamide, or LSD for short. When he’d originally stumbled upon it inadvertently, he’d dismissed it as having no obvious use—it wasn’t clear that it did anything. But in what he would later describe as a moment of intuition, one day he felt instinctively that he ought to submit the compound to more thorough testing, in case it had some useful properties he’d initially overlooked.
On that Friday in spring, Hoffman was forced to leave work early when he began feeling restless and dizzy; and when he got home, he sank into a stupor he described as similar to being drunk. Finding daylight unpleasant at the time, he closed his eyes and drifted into a dreamlike state in which he experienced a play of shapes, colors, and images that ran through his mind uninterrupted for about two hours.
Unable to come up with any other explanation for his bizarre experience, he decided that, in spite of his usually meticulous laboratory manner, he must have inadvertently gotten a trace amount of the mystery compound on his skin. Wanting to know if it was indeed this strange new chemical that had caused his Friday afternoon to take such an unusual turn, he decided to take a risk: When he returned to work on Monday, he deliberately ingested LSD.
At 4:20 that afternoon he mixed what he thought was a small sample of LSD in water (about a quarter of a milligram), and drank it. At 5pm, he wrote in his journal:
Beginning dizziness, feeling of anxiety, visual distortions, symptoms of paralysis, desire to laugh.
Those were the last words he could write that day, and even that was only through an immense effort of will. His perceptions of reality were drastically altered, and he struggled to speak, able only with great difficulty to communicate to his lab assistant that he needed to go home.
At this point, the drug’s effect on him had become severe enough to be frightening. Objects in his surroundings took on grotesque shapes and seemed imbued with malevolent life. His sense of his self—both his bodily awareness and his concept of will—were disturbed as well. At times he thought he’d been possessed by a demon; later, he wondered if he was dying, or perhaps had died already. He lamented the thought of leaving behind a wife and three young children, and despaired at the dark irony that his work should be cut short by the very substance he’d discovered.
When he awoke the next morning, he felt no unpleasant after-effects, apart from still feeling a bit tired. He wrote, “A sensation of well-being and renewed life flowed through me. Breakfast tasted delicious and gave me extraordinary pleasure. When I later walked out into the garden, in which the sun shone now after a spring rain, everything glistened and sparkled in a fresh light. The world was as if newly created.”
Full thesis here: http://dspace.mit.edu/handle/1721.1/76174
Dont’ Call it a Seagull
Next to me on the pier, another birder was surveying the flock with a far more discerning eye
than my own. Jeremiah Trimble, a lanky, sandy-haired gull fanatic in his early thirties, has been a birder ever since he was a kid growing up on Cape Cod. For the past ten years he has also been the Curatorial Associate of the Ornithology Department at the Harvard Museum of Comparative Zoology. Every weekend in the winter, Trimble embarks on the fifty-minute drive from Cambridge to Gloucester to make the rounds at all of the best gull-watching spots—Jodrey Pier, Niles Pond, Eastern Point, the Elks Lodge.
Trimble especially hopes for vagrant birds that have been blown far off course while migrating. That’s another reason for watching gulls in the winter: the tantalizing possibility of seeing a real rarity. Just last week, he got photographs of a Slaty-backed Gull, a bird normally found between Japan and Alaska. The same individual, or one that looks identical in photos, was originally spotted a few days before that at a dump in Maine—the very first record of a Slaty-backed Gull in the state. (It wasn’t a first for Massachusetts, but it was still the talk of the Massbirds listserve for days.)
The average bystander would notice nothing exceptional about a Slaty-backed Gull. To me, it might have passed for a Great Black-backed Gull or even an extra-dark Herring Gull. But for someone like Trimble, its subtly different plumage—the sooty smudge around its eye, the white “string-of-pearls” spots on its black wingtips as it stretched its wings—stood out like a neon sign. Even though the gull was surrounded by other birds, he picked it out immediately as a third-year Slaty-backed.
That’s right: Trimble not only knows what the adults of the different species look like, both in breeding season and in winter, but he knows the different plumages for first-year birds, second-year birds, and third-year birds. He can identify transition plumages while birds are molting. He can even tell when two gull species seem to have hybridized and produced offspring with roughly intermediate characteristics.
Birders like Trimble are hooked on identification. Some of them see it as an intellectual challenge, a puzzle to be solved like the Sunday crossword. Some particularly enjoy the thrill of the chase, the camaraderie of like-minded people, the excuse to be outside. But on a fundamental level, birders are simply acting on an impulse that’s familiar to everyone: we all want to know who’s who. We want names and groups to hold onto, so we can begin to have relationships with our fellow inhabitants of earth.
My botany professor in college once confided to me how deeply she loves recognizing plants everywhere she goes. “It’s like seeing friends!” she said. When you recognize the identity of another living being, you become more connected to it and the world you both live in. I think that’s why, underneath everything else, humans are so drawn to identification. That’s why we go through the world persistently asking who are you?
There are different ways of answering that question. For some people, it’s enough to be able to tell a bird from a fish. For birders, it’s often about pinpointing identity as finely as possible, down to species or even subspecies. And for the science-minded, there’s a whole other layer to the question: What makes you who you are?
Full thesis here: http://dspace.mit.edu/handle/1721.1/76140
May 15, 1983. A man weighing 217 lbs and 6 ft 2 inches in height throws an 800 gram hollow metal spear. He throws from a run and it takes him 18 steps to do so, 16 plus 2 for the release, using 86 ft out of the 110 ft allowed. He draws back the spear at exactly 46 ft. He knows all of this without thinking about it because he has practiced this same run-up hundreds of thousands of times since the late 1970s.
Just before the release, the man sticks his left leg out to vault over it, putting over 1000 lb per square inch on his block foot, half the jaw force of a crocodile clap. The spear is released at 72 miles per hour, rotating counterclockwise at 1 rotation per second and 14 degrees parallel to the ground. We know this because even more significant than the throw itself was the way it was captured; with high speed video, at 200 frames per second. This technique would revolutionize the way track and field is studied following a publication of the throw’s analysis in the Journal of Applied Biomechanics. Since then, high speed video in sports has given us an understanding of what our bodies do that has caused us to adapt how we train, akin to what a series of photographs of a galloping horse called “The Horse in Motion” did for a how a horse moves in 1878.
Yet the first high speed video of a biomechanical movement captured something that had also never happened before. At the 1983 Pepsi Invitational Track Meet in UCLA, Tom Petranoff had set the new world record for the javelin throw by just under ten feet, the height of a basketball hoop (exactly 3.00 meters). 327 ft later – further than the length of a football field – Tom Petranoff’s javelin lands in the grass, head first. We know this because it was measured.
Throwing the javelin is conceptually simple but physically complex. A long throw looks effortless and fluid on film, yet this observation is literally only skin deep. For all that has been written and recorded about our bodies, we still know little about how we enable ourselves to do a physical motion like throwing. A throw involves millions of muscle parts all timed by an unconscious process that is not understood, making the javelin throw arguably one of the most technically complex athletic events in track and field.
And while some animals may run faster or jump higher than us, no other animal comes close to doing this. Throwing on the run is uniquely human. We share 98% of our genetic material with chimpanzees, so much so that Jared Diamond’s book on human origins was called “The Third Chimpanzee”, but an alien scientist who landed on Earth would know immediately from our physical abilities that we are a totally different animal.
It is well known that humans have unique cognitive abilities, what we call intelligence. Lesser known are our unique physical abilities, such as throwing. The question naturally arises: is there a causal link and if so, how would we know? To explain these behavioral traits and what separates us from other animals in form and function, many anthropologists have spent much time theorizing on how events in history could have fortuitously fallen into place to get us physiologically and cognitively to where we are today.
Full thesis here: http://dspace.mit.edu/handle/1721.1/76142
There is a mischievous twinkle in Paul Schechter’s eyes as he reaches from thefloor and pulls up a box, big enough to fit a large bottle of wine. According to the label, it contains a small amateur telescope. Schechter places it on the table in front of him and says, “That’s a smaller telescope!” before laughing at his own joke and apologizing.
Schechter, an astrophysicist at MIT, had just been asked about astronomy projects that are smaller and more affordable than the trend of “large” telescopes and the even larger “extremely large” telescopes. The small boxed telescope, fit for the backyard, is his response-in-jest.
He is more serious in his actual answer: “There’s a telescope called the OGLE telescope. It’s a 1.3-meter telescope, and those guys have done fantastically good stuff by choosing their problems very, very carefully, even though they don’t have the light- gathering power of bigger telescopes.” Then he adds with another laugh, “It’s possible to do great science with smaller telescopes, but it’s much easier with big telescopes!”
OGLE, which stands for Optical Gravitational Lensing Experiment, is not shrt of stellar achievements, including contributing data that revealed that our galaxy contains billions of exoplanets. However, its ability is definitely limited: OGLE astronomers mostly ogle at the Milky Way and the Magellanic clusters (two satellite galaxies of the Milky Way), a move calculated to take advantage of those regions’ abundances of close-by and closely-spaced stars.
Logically, astronomers would not mind having telescopes with more power and flexibility. After all, they have a lot of questions.
Once the trend of six to ten meter telescopes became well-established in the 1990s, proposals for the extremely large telescopes, or ELTs, began popping up, each boasting designs with aperture widths on the order of twenty, thirty, fifty, and even a hundred meters. Emails were exchanged, phone calls were made, conferences were held, and ideas flew. “I think I once counted eighteen different sorts of future giant telescope proposals that were out there,” says Larry Stepp, an engineer who has worked on ELT designs for over a decade. “And over time, these began to coalesce into more real projects.”
The ELTs hold a lot of promise: they will be able to collect images more quickly than their existing counterparts, and look into more distant reaches of the universe. Their images will be ten times sharper than that of the Hubble Space Telescope, despite being on the ground rather than in space. Their sheer sizes make them too big to launch into orbit, but there are clever technological workarounds that negate the atmospheric turbulence that would otherwise muddle the view of the sky. Armed with an ELT, one could easily spot a truck on the moon, though of course, scientists would use the instruments for research far more ambitious. These telescopes, once built, promise astronomical delights, such as snapping pictures of exoplanets, studying the universe’s very first stars, exploring the physics of black holes, and investigating the natures of dark energy and dark matter—the most pressing questions in astronomy today.
Full thesis here: http://dspace.mit.edu/handle/1721.1/77472