Last night, I headed down to the beautiful Bell House, in the Gowanus section of Brooklyn, for this month's Secret Science Club lecture featuring Dr Hod Lipson, profession of mechanical engineering and director of the Creative Machines Lab at Columbia University. The subject of Dr Lipson's lecture was AI and robotics.
Dr Lipson began his lecture by declaring that, after two decades of research, the fields of artificial intelligence and robotics took off in the last couple of years. He posited the questions, what are the trends in robotics, where are they going, and what are the 'game changers' in the field? He urged the audience to think long term, then, displaying a movie still, intoned that 'a long time ago and many galaxies away', the conventional wisdom was that robots would take humanoid forms. Also, there was a 'lone genius' model of scientific research which is not the norm.
Today, there are millions of robots, mainly in factories. These robots are powerful and precise, but they are not clever. Dr Lipson posed the question, 'how do you make robots more adaptive?' He noted, regarding artificial intelligence, that there are 'tidbits' everywhere, but not in the manner in which most laypersons think- AI is involved in investment banking, weather prediction, and music programming. He joked that artificial intelligence paired with real money. The number of robots is growing and the role of artificial intelligence in daily life is growing- what had been talked about for decades is now common.
Dr Lipson then displayed a picture of a Roomba and then a picture of a drone, then quipped that computers that once would have netted a researcher a PhD are now used as toys. He then displayed a quote by Marc Andreessen: "Software is eating the world." Artificial intelligence is now 'infusing' robots. Dr Lipson then showed us a video of table tennis champion Timo Boll playing a match against a robot dubbed KUKA:
In the match, the robot dominated until Mr Boll was able to figure out ways to thwart it- robots perform well when things are in the right place at the right time, but they have trouble coping with 'corner cases'. It is difficult for robots to adapt to conditions that their programmers didn't anticipate.
Dr Lipson then backtracked a bit, giving a quick history of robotics. In 1912, John Hammond, Jr. and Benjamin Miessner designed a self-orienting robot which was programmed to turn towards light sources. This robot, designed to deliver explosives behind enemy lines, was described as being able “to inherit almost superhuman intelligence.” Throughout the 20th century, the goal was to develop exponentially more powerful technologies, a goal expressed most succinctly in Moore's Law. The exponential growth of processing power is a small part of the equation... something was missing in the hardware and software, and a political war arose within the computer science community about how to achieve artificial intelligence. Should programming be a 'top down' process or should computer scientists design computers that learn? In traditional computer programming, a programmer writes code and the computer uses the algorithm to solve a problem. With the advent of artificial learning, the computer learns how to solve a problem. The problem with top-down programming is that it is impossible to think of every 'if then else' statement needed to cope with a multitude of conditions. The top down approach doesn't work with all problems- there's no code to keep a driverless car on the road if the 'if then else's' are insufficient.
In 1957, Frank Rosenblatt of Cornell University developed a neural network that could distinguish simple shapes, and pioneered the field of machine learning. One early development in machine learning involved a computer that could play checkers. The original programming entailed writing algorithms instructing the computer to take opponent's pieces- the computer played a mediocre checkers game because humans made unexpected moves, things that the programming didn't anticipate. The programmers then changed their strategy- they collected data and programmed the computer to 'play' many games of checkers, mimicking the winning games. The computer was able to accumulate 'lifetimes' of checker playing experience. Then came the development of chess playing computers, also being able to learn multiple chess games until able to beat grandmasters. Recently, a computer was able to beat a champion Go player. Dr Lipson asked if this was the end of an era, then said that computers are poised to tackle the real world.
The subject of the lecture then shifted to the DARPA autonomous vehicle challenge. Dr Lipson referred to the work of Stanford's Sebastian Thrun, who addressed the quandry, does one use machine learning or big data? Does one teach an autonomous vehicle to drive like one would teach a human how to drive? Ultimately, the machine learning approach was taken. In 2007, there was a collision between the Cornell driverless car and the MIT driverless car, which Dr Lipson was quick to blame on the MIT team- the Cornell driverless car interpreted an immovable object as another vehicle then stopped in order to give it right-of-way, the MIT vehicle passed the stationary Cornell vehicle just as it reinterpreted the immovable object and started to steer around it. Both machines 'sensed' the world around them, but they didn't 'understand' what they sensed. The eyes worked, but they could not see. There was a failure of understanding.
Along with artificial intelligence, advancements in material science are also revolutionizing robotics. New synthetic 'muscle fiber' has been developed and 3D printers can make pieces which mimic organic structures. The programming of these new robots marries algorithms with machine learning, combining to produce an explosion in AI. In the area of Big Data, the number of cameras has increased exponentially... visual data, once difficult to obtain, is now ubiquitous. While academia always had access to 'fast' computers, this data was the missing element in early artificial intelligence efforts.
In 2012, ImageNet, originally a set of one million images, was created to test artificial intelligence. A million images were classified in thousands of categories, and the AI's were tasked to label novel images correctly. Dr Lipson jocularly referred to ImageNet as a 'Mechanical Turk' for recognizing images. Describing the project, Dr Lipson noted that it was 'freelancers training AI to take their jobs'. Originally, the ImageNet had a 25% error rate, a rate altogether unacceptable to employ in driverless cars. In 2012, a team from the University of Toronto developed an algorithm called SuperVision which dropped the error rate to 15%. Dr Lipson joked that this dramatic improvement was 'like seeing Jesus'. The real game-changer in visual recognition is the fact that the code is open source. While Frank Rosenblatt's early neural networks involved many 'layers' of wires, the new neural networks for visual recognition involve many 'layers' of code- it's a 'souped up' version of the neural network. Humans typically have an error rate of 5% in ImageNet, much of the error rate is due to an inability to distinguish images due to unfamiliarity with, say, breeds of dogs, or types of lizards. In 2015, a Microsoft team was able to bring the error rate down to 3%, for the first time in history, machines were able to 'understand' images better than humans were.
Another game changer in AI is the improvement of voice recognition and a refinement of image understanding- for example, the ability for a computer to recognize a 'kiss' in a movie. Computers are gradually understanding lots of images and the connections between objects.
Returning to the subject of driverless cars, Dr Lipson noted that it's difficult to distinguish things on the road... there are many 'corner cases'. Is an object a pothole? An oil spill? A shadow? The object recognition abilities of AI's couldn't be trusted. Is an object a fire hyrant, or a kid? With new hardware and new software bringing better image recognition ability to machines, the last link in the 'driverless car' puzzle has been made.
Dr Lipson then noted that AI has wormed its way into all aspects of our lives. Because the algorithms are open-source, anyone can use them. There is a 'dark side', though, we don't always understand how the algorithms work. In one particular instance, an image recognition program 'learned' how to distinguish faces, even though it was never programmed to do so- the AI learned that tracking human faces was useful. While PhD candidates were working on facial recognition software, this particular computer just learned it. With the advent of the 'cloud', Dr Lipson joked, what one robot learns, all robots know. With shared data, a driverless car can draw upon thousands of lifetimes of driving, a robot doctor can 'learn' on millions of patients.
Dr Lipson opined that Isaac Asimov's Three Laws of Robotics were garbage- we don't know how robots learn and make decisions. The best way to track AI is to use AI to do so. The goal is to make 'curious and creative mechanisms'. Most artificial intelligences are designed to make decisions- buy or sell? These AI's take in information and make decisions.
Another kind of intelligence is creativity- humans create, but synthesis, unlike analysis, is difficult to teach. In developing 'creative' robots, the best inspirations are not from humans, but from evolution. In one particular case, a 3D printed robot was designed using a biological model as inspiration, and an 'evolutionary' algorithm was used develop a 'crawling' robot. Originally, the robot didn't have a self-image, and didn't know how to walk. First the robot had to discover what it looks like, and initially it flailed around until a self-image developed. Then the robot learned that it had four legs and used this information to develop a method of walking. The robot had to learn to walk, Dr Lipson joked that it was a far cry from an 'evil spider robot'. Then, when the robot had learned how to walk, one leg was removed, so the robot had to modify its self image and adjust its locomotion. Other gaits were generated for different damaged morphologies... here's a video of Dr Lipson explaining the robot's learning process:
The subject of the talk then shifted to artistic ability- Dr Lipson quipped that human beings have dreams, even dogs have dreams, but can a computer paint real 'art' or write a symphony? He described efforts to teach a robot how to paint, and noted that robots can paint decent portraits. In another instance, he described an AI that created an image of a double pendulum in action, and the AI was able to duplicate the equation describing the pendulum's motion using only algebra.
The lecture then veered into the subject of metacognition, thinking about thinking. Dr Lipson joked that computer scientists can't mention cognition or consciousness until they get tenure. Can one speak of 'robopsychology' or theory of mind? In the case of the walking robot, a self image had to develop, and as alterations were made to the machine, the self-image needed changing. Currently, artificial intelligences have no theory of mind- for example, drones aren't aware of other drones... yet. Dr Lipson brought up the subject of AI affecting jobs, and noted that the real question is 'what will people do to other people with artificial intelligences?' Utopia or dystopia? The answer will rely on human use of AI. Today, we are at a cusp, with a Cambrian explosion of robotics on the horizon. The main evolutionary changes in the Cambrian explosion involved the evolution of eyes, the ability to perceive the world, in many animal lineages. There is an explosion of AI 'forms' and successful robots will be 'rewarded' with replication in an echo of the evolutionary model.
In the Q&A, some bastard in the audience asked about the 'holy grail' of robotics, the development of self-replicating machines such as the hypothetical Von Neumann probe. While noting that the crude mechanics of self-replicating machines are challenging, software self-replication is possible, dependent on providing building blocks and definitions of self-replication. Another question, regarding robot consciousness elicited the joke that that's a topic restricted to tenured professors- Dr Lipson then asked, is a dog conscious? Consciousness doesn't have to achieve the level of human consciousness- are self-awareness and self-simulation sufficient? Another question involved biological models for robotics, and Dr Lipson cited evolutionary theory and neuroscience as useful complemetary disciplines. Regarding robotic medicine, how would one ensure ethical behavior? While one cannot program ethics into a computer, an AI could be taught as a child is taught- give examples and hope for the right choices. Another question involved maintaining control over AI's- with machines being able to exceed human limitations, perceiving more wavelengths, more frames per second, there is less human control over those machines. Regarding the open-source nature of much of computer science, the open-source movement has led to more data and better algorithms, but while corporations develop open source algorithms, they tend to keep the data private. The final question regarded driverless vehicles... is vehicle to vehicle 'awareness' better than a totally autonomous system? Dr Lipson preferred totally autonomous vehicles- they are harder to hack and no transponders are needed.
Once again, the Secret Science Club served up a fantastic lecture. Kudos to Dr Lipson, Dorian and Margaret, and the staff of the beautiful Bell House. As an added bonus, last night my friend Peter, originally of Yonkers but currently residing in San Diego, was able to attend the lecture. He's involved in app development and he had good things to say about the lecture. It's nice to see a nerd's nerd reacting to the lecture and the vibe- it's not every day that one attends a hard science lecture that appeals to both experts and laypersons accompanied by a fully stocked bar... it's just every Secret Science Club day.
Showing posts sorted by relevance for query robot chess. Sort by date Show all posts
Showing posts sorted by relevance for query robot chess. Sort by date Show all posts
Tuesday, June 14, 2016
Wednesday, August 21, 2013
Secret Science Club Post Lecture Recap: This Indecision's Bugging Me
Last night, I headed to the beautiful Bell House in the Gowanus section of Brooklyn for the latest Secret Science Club lecture, featuring neurologist Dr Anne Churchland of the Cold Spring Harbor laboratories. Dr Churchland seeks to bridge the gap between our knowledge of sensorimotor reflexes and more complex behavior. Specifically, she studies the decision-making process.
Since time immemorial, humans have wondered what makes us who we are and what makes us decide to do what we do. Disease states and drugged states played some role in elucidating the mysteries of the brain, but the brain was largely unknowable. Now that we have the ability to measure neuronal activity, and to perturb neuron groups, we can begin to understand the neural circuits which give rise to behavior. The ability to understand the brain's "circuitry" can also help us solve clinical problems, such as depression, which has a major cost. Approximately one in seventeen individuals in the U.S. suffers from severe mental illness.
Dr Churchland covered the topic of reflexive behavior, ranging from such simple, easily understood behaviors as the knee-jerk reflex to the more complex vestibulo ocular reflex, which stabilizes images when one moves one's head. Dr Churchland ran us through some tests to demonstrate the vestibulo-ocular reflex, having us move our fingers in front of our eyes and asking us if the image were blurred (she joked about how the beer-guzzling audience would naturally see blurred images).
She then showed us some headlines from the popular press which made a butchery of neuroscience, including the embarrassing NY Times Headline "You Love Your iPhone. Literally.", which cited activation of the insular cortex as evidence of "love" when one handled one's iPhone... though the insular cortex plays a role in disgust responses as well as love.
Dr Churchland posed the question, "Why is understanding the brain so hard?" One reason for this difficulty is the complexity of the brain- the brain is composed of eighty-six billion neurons, which form various structures within the brain. Another difficulty is that introspections about brain function are misleading. Dr Churchland used the game of chess as an example of incorrect intuitions about brain functions... while most laypersons believe that understanding the various chess moves is more difficult than understanding how the actual pieces are physically moved- in 1997, a computer was able to outplay a chess grandmaster , but the development of artificial intelligences which can manipulate objects is in its infancy (Dr Churchland showed a tragicomic video of a robot hand trying to pick up a coffee cup). Additionally, brains do different things in different animals. Dr Churchland cited bats and mice as two mammals with very different brains. One can look to see how problems are solved in animals- this works well for kidneys and hearts, which are very similar in different animals, but brains are much more diverse. A model system has to be chosen carefully due to the evolution of brains to serve different needs for different ecological niches.
Dr Churchland's major goal is to bridge the gap between sensorimotor reflexes and complex behaviors. Reflexes are "locked in" timewise- when one's knee is tapped with a mallet, the reflex occurs in an inflexible timeline. Complex decisions take place on a more flexible time scale, and integrate many systems of sensory input. Dr Churchland cited the purchase of a car as an example- one uses various sensory inputs to evaluate a vehicle, and there is no stereotypical time scale on which a final decision is made.
Dr Churchland went on to demostrate multisensory integration, and the differences between our processing of stimuli. Auditory stimuli are not very good in a spatial sense- it is often hard to pinpoint where a sound is coming from. Visual stimuli are very good at spatial resolution. As an example of this disparity, Dr Churchland cited a ventriloquist, who is able to exploit our poor auditory special recognition by providing a deceptive visual stimulus- we are "fooled" by the dummy's mouth moving, so it seems like the sound is emanating from the dummy. As a more glaring example, she pointed to the speakers around the room, from which her voice was emanating, and noted that none of us had any problem perceiving her as the source of her voice.
The visual system is not particularly good at timing, while the auditory system is much better at it- by flashing a dot once, but sending out two "beeps" in rapid succession, she was able to "fool" the audience into believing that two dots were flashed.
Subjects weigh incoming information according to its reliability- Dr Churchland cited a 2002 paper by Ernst and Banks (PDF) which dealt with the integration of visual and haptic (touch) stimuli- when deceptive visual stimuli are presented, a subject can use touch to correct perception. The brain can change from one stimulus to another moment-to-moment, with regards to environmental statistics- multisensory integration is geared towards a statistically optimal condition.
In her lab, Dr Churchland uses rodents as subjects- rodents are able to judge stimuli rates to be high or low. A subject was confronted by an array of LEDs and speakers and would respond to the auditory and visual stimuli in order to gain a reward, a drink of water. A low stimuli rate would indicate that the rat would be rewarded by placing its snout in the left-hand dispenser, a high stimuli rate would indicate the right-hand dispenser. Twelve "events" per second was chosen as the arbitrary "cutoff" between high and low rates. While the task that had to be performed by the rat was unnatural, it nevertheless tapped into natural neural pathways. A rat could be trained to shape its decision making behavior (at this point, Dr Churchland showed a hilarious "training" montage accompanied by the song The Eye of the Tiger). After showing video of the rat performing the necessary tasks, Dr Churchland called for a volunteer from the audience, and a game fellow named Issac was brought up to the stage to undergo the same "test" as the rats. He scored a whopping 85% success rate, better than any rat. For his efforts, Dr Churchland awarded him with a packet of string cheese.
Of course, studying behavior is only the first step in the "journey"- imaging of the brain was conducted while the rat was engaged in the decision making process. To supplement brain imagine, tetrodes fifteen microns in diameter were placed in the rats' skulls to measure neural activity. There was evidence of neurons firing in the posterior parietal cortex, which is imaged in this video, which indicates that the PPC plays a role in decision making.
In the Q&A some bastard in the audience asked Dr Churchland how the rats dealt with conflicting stimuli, say a high rate visual stimulus combined with a low rate auditory stimulus. She indicated that, in the case of conflict stimuli, the more reliable stimulus was the basis of the decision- a bright visual stimulus would be weighted more heavily than a low auditory stimulus. Another questioner in the audience asked about the role of glia in the brain, and Dr Churchland indicated that the glia play a "scaffolding" role- they help to define the structure of the brain.
Once again, last night's lecture was another triumph of the Secret Science Club. Dr Churchland gave a great presentation about the workaday aspects of a scientist's research, and gave a wonderful view into the efforts to increase our knowledge about the "divide" between simple reflexive behavior and complex behavior. For a brief taste of the subject of the lecture, here's a video of Dr Churchland discussing research that she had conducted using primates as subjects:
Since time immemorial, humans have wondered what makes us who we are and what makes us decide to do what we do. Disease states and drugged states played some role in elucidating the mysteries of the brain, but the brain was largely unknowable. Now that we have the ability to measure neuronal activity, and to perturb neuron groups, we can begin to understand the neural circuits which give rise to behavior. The ability to understand the brain's "circuitry" can also help us solve clinical problems, such as depression, which has a major cost. Approximately one in seventeen individuals in the U.S. suffers from severe mental illness.
Dr Churchland covered the topic of reflexive behavior, ranging from such simple, easily understood behaviors as the knee-jerk reflex to the more complex vestibulo ocular reflex, which stabilizes images when one moves one's head. Dr Churchland ran us through some tests to demonstrate the vestibulo-ocular reflex, having us move our fingers in front of our eyes and asking us if the image were blurred (she joked about how the beer-guzzling audience would naturally see blurred images).
She then showed us some headlines from the popular press which made a butchery of neuroscience, including the embarrassing NY Times Headline "You Love Your iPhone. Literally.", which cited activation of the insular cortex as evidence of "love" when one handled one's iPhone... though the insular cortex plays a role in disgust responses as well as love.
Dr Churchland posed the question, "Why is understanding the brain so hard?" One reason for this difficulty is the complexity of the brain- the brain is composed of eighty-six billion neurons, which form various structures within the brain. Another difficulty is that introspections about brain function are misleading. Dr Churchland used the game of chess as an example of incorrect intuitions about brain functions... while most laypersons believe that understanding the various chess moves is more difficult than understanding how the actual pieces are physically moved- in 1997, a computer was able to outplay a chess grandmaster , but the development of artificial intelligences which can manipulate objects is in its infancy (Dr Churchland showed a tragicomic video of a robot hand trying to pick up a coffee cup). Additionally, brains do different things in different animals. Dr Churchland cited bats and mice as two mammals with very different brains. One can look to see how problems are solved in animals- this works well for kidneys and hearts, which are very similar in different animals, but brains are much more diverse. A model system has to be chosen carefully due to the evolution of brains to serve different needs for different ecological niches.
Dr Churchland's major goal is to bridge the gap between sensorimotor reflexes and complex behaviors. Reflexes are "locked in" timewise- when one's knee is tapped with a mallet, the reflex occurs in an inflexible timeline. Complex decisions take place on a more flexible time scale, and integrate many systems of sensory input. Dr Churchland cited the purchase of a car as an example- one uses various sensory inputs to evaluate a vehicle, and there is no stereotypical time scale on which a final decision is made.
Dr Churchland went on to demostrate multisensory integration, and the differences between our processing of stimuli. Auditory stimuli are not very good in a spatial sense- it is often hard to pinpoint where a sound is coming from. Visual stimuli are very good at spatial resolution. As an example of this disparity, Dr Churchland cited a ventriloquist, who is able to exploit our poor auditory special recognition by providing a deceptive visual stimulus- we are "fooled" by the dummy's mouth moving, so it seems like the sound is emanating from the dummy. As a more glaring example, she pointed to the speakers around the room, from which her voice was emanating, and noted that none of us had any problem perceiving her as the source of her voice.
The visual system is not particularly good at timing, while the auditory system is much better at it- by flashing a dot once, but sending out two "beeps" in rapid succession, she was able to "fool" the audience into believing that two dots were flashed.
Subjects weigh incoming information according to its reliability- Dr Churchland cited a 2002 paper by Ernst and Banks (PDF) which dealt with the integration of visual and haptic (touch) stimuli- when deceptive visual stimuli are presented, a subject can use touch to correct perception. The brain can change from one stimulus to another moment-to-moment, with regards to environmental statistics- multisensory integration is geared towards a statistically optimal condition.
In her lab, Dr Churchland uses rodents as subjects- rodents are able to judge stimuli rates to be high or low. A subject was confronted by an array of LEDs and speakers and would respond to the auditory and visual stimuli in order to gain a reward, a drink of water. A low stimuli rate would indicate that the rat would be rewarded by placing its snout in the left-hand dispenser, a high stimuli rate would indicate the right-hand dispenser. Twelve "events" per second was chosen as the arbitrary "cutoff" between high and low rates. While the task that had to be performed by the rat was unnatural, it nevertheless tapped into natural neural pathways. A rat could be trained to shape its decision making behavior (at this point, Dr Churchland showed a hilarious "training" montage accompanied by the song The Eye of the Tiger). After showing video of the rat performing the necessary tasks, Dr Churchland called for a volunteer from the audience, and a game fellow named Issac was brought up to the stage to undergo the same "test" as the rats. He scored a whopping 85% success rate, better than any rat. For his efforts, Dr Churchland awarded him with a packet of string cheese.
Of course, studying behavior is only the first step in the "journey"- imaging of the brain was conducted while the rat was engaged in the decision making process. To supplement brain imagine, tetrodes fifteen microns in diameter were placed in the rats' skulls to measure neural activity. There was evidence of neurons firing in the posterior parietal cortex, which is imaged in this video, which indicates that the PPC plays a role in decision making.
In the Q&A some bastard in the audience asked Dr Churchland how the rats dealt with conflicting stimuli, say a high rate visual stimulus combined with a low rate auditory stimulus. She indicated that, in the case of conflict stimuli, the more reliable stimulus was the basis of the decision- a bright visual stimulus would be weighted more heavily than a low auditory stimulus. Another questioner in the audience asked about the role of glia in the brain, and Dr Churchland indicated that the glia play a "scaffolding" role- they help to define the structure of the brain.
Once again, last night's lecture was another triumph of the Secret Science Club. Dr Churchland gave a great presentation about the workaday aspects of a scientist's research, and gave a wonderful view into the efforts to increase our knowledge about the "divide" between simple reflexive behavior and complex behavior. For a brief taste of the subject of the lecture, here's a video of Dr Churchland discussing research that she had conducted using primates as subjects:
Wednesday, June 16, 2021
Secret Science Club Zoom Lecture: Shape
Tonight, my great and good friends at the Secret Science Club hosted a Zoom lecture featuring mathemetician Dr Jordan Elleberg of the University of Wisconsin, Madison. Dr Ellenberg's newly published book is Shape: The Hidden Geometry of Information, Biology, Strategy, Democracy, and Everything Else.
Dr Ellebberg began his lecture by noting that everything is connected, showing a 'map' of connections of topics in his book. Ronald Ross was a physician who determined that malaria was transmitted by mosquitos. Ross was an indifferent doctor, but had a love for mathematics, and he applied mathematical models to epidemiology. Ross wanted to formulate a theory of phenomena, starting with epidemics. His work in this field was the beginning of mathematical modeling. He applied it to the problem of malaria... eliminating malaria would involve eliminating mosquitos, which is impossible. Mosquitos can be temporarily eliminated from an area- how long would it take for them to repopulate an area. Mosquitos do not move in predetermined fashion, they move largely at random. Ross enlisted mathemetician Karl Pearson to couch a model of mosquito repopulation of an area in neutral terms, removing references to insects- The Problem of the Random Walk.
Botanist Robert Brown noticed the movement of particles in a medium, which became known as Brownian motion- he wondered if is it a vital life principle, noticing it in pollen first. Brown tested it on organic and non-organic materials (including a 'fragment of the Sphinx'). Albert Einstein noted that the molecules are colliding, causing this movement. This motion can be figured only on a basis of probability.
Russian mathematician Andrey Markov had a reputation for being furious- he was angry that Tolstoy was excommunicated while he was not, so he ended up being excommunicated as well. He approached the problem of the Law of Large Numbers, which basically states that if one were to flip a coin numerous times, the more times it is flipped, the probability of heads and tails approach fifty percent increases. Flip a coin ten times, there is a good chance there will be six heads and four tails... flipping one thousand coins, having six hundred heads and four hundred tails would be less probable. Markov formulated the concept of the Markov Chain. He applied the Markov chain to determine the sequence of vowels and consonants in Pushkin's poem Eugene Onegin
Dr Ellengram then played around with bigrams- what letters are likely to follow other letters? He mentioned playing with a computer game called AI Dungeon which can be used to generate texts. He presented an artificial intelligence generated text about geometry- not quite convincing, but with an occasional flash of brilliance such as: "But squares aren't just shapes, they're also numbers!" Can machines replace humans? There is a line of difficulty from, say Tic Tac To to a perfect Go game- computers aren't smarter if they can beat humans at chess or Go, it's a one dimensional difficulty issue... a robot may beat a human at chess, but it can't fold a shirt. Machines will be great collaborators for us- we must determine which tasks they can outperform us in. Dr Ellenberg hopes they can be capable partners.
The lecture was followed by a Q&A session, which began with a question about gerrymandering- new districts are going to be drawn, this gives the people who draw these lines great power over who gets elected. In Wisconsin, the current legislators draw the maps, which is a problem. Legislators are given the keys to thwarting the electorate. Districting is a geometric problem- there are districts which look like 'polyamorous octopuses'. Mathematical tests can determine how bad gerrymandering is.
How many holes in a straw? It depends.
Dr Ellenberg criticized a mathematical approach which separates the subject into discrete courses of study- mathematical fields are connected.
The Random Walk is a probability problem, but also a geometry problem. Dr Ellenberg sees most math as having a geometric component.
Was Lewis Carroll aware of Bigrams? Jabberwocky seems to hint that he was... his fake words sound plausible, but Dr Ellenberg wasn't sure if he were aware of bigrams... it would be a great fake theory to promulgate, Dr Ellenberg joked.
Squaring the circle- problem for the ancient Greeks, could a square be created with the same area as a circle? It became a symbol of a difficult problem. Lincoln, a geometry enthusiast, used this metaphor to express difficulty.
Mathematics is built one the one hand on rigid reasoning and on the other hand on intuition. Geometry is based on our bodies, our two-dimensional field of perception and our three-dimensional space.
Regarding internet searches, the search engines use a random walk process to determine the priority of search results.
Regarding the use of math to map pandemics, Dr Ellenberg referred back to Ross attempt to formulate a theory of phenomena. People move, pathogens are transmitted, this is a geometric problem.
Once again, the Secret Science Club has dished out another fantastic lecture, a humorous deep dive into esoteric topics. Kudos to Dr Ellenberg, Margaret and Dorian. For a small taste of the Secret Science Club experience, here is the Good Doctor speaking on the subject of his new book:
Pour yourself a nice beverage, sit back, and soak in that SCIENCE!!!
Wednesday, September 18, 2013
Secret Science Club Post Lecture Recap: Of Two Minds
Last night, I headed down to the beautiful Bell House in the Gowanus section of Brooklyn for the latest Secret Science Club lecture, featuring Dr Moran Cerf, who is one of those scientist/rockstar figures, a former hacker/security expert and Moth "Grand Slam" story winner.
Dr Cerf's lecture dealt with decision making and the brain/body interface. Why do individuals want one thing, but do the opposite thing? At times, it seems that an individual is two people fighting for dominance in one body. How does the brain interact with the body in everyday life?
Dr Cerf illustrated this conundrum by relating the tragic tale of Charles Whitman, the infamous University of Texas "tower sniper". After the mass shooting, the police tried to determine a cause for the shooting spree. Whitman was generally seen as a quiet, law abiding man, but his diary told an alarming tale. Whitman wrote that he "feels like he's not the same guy" and described having violent urges. Before he embarked on his murderous foray, he indicated that he wanted an autopsy performed on him and left a check to pay for it. When an autopsy was performed, a tumor the size of a walnut was found in his brain which impinged on his amygdala.
I another instance, a patient with a frontal lobe tumor began to exhibit signs of sexual deviancy, including pedophila. The removal of the tumor caused him to revert back to his previous sexual mores, but a partial recurrence of the tumor caused a return of the symptoms until it too was removed.
To illustrate the ability of individuals to "color" their perceptions of events, Dr Cerf cited a study (PDF) in which which a group of colonoscopy patients were told to rate the pain they experienced on a ten-point pain scale. Half of the patients experienced an interval at the end of the procedure in which the tip of the scope remained in their rectums. Oddly enough, the patients with the extended colonoscopy rated their experience as less painful than the patients whose colonoscopies were shorter in duration. It is probable that they experienced fewer "highs and lows" during the procedure than those with a quicker colonoscopy, and that the patients with the quicker procedure assessed it from moment to moment, rather than assessing it as one single event.
Studying the "competing influences" in the brain poses some ethical dilemmas- while studies using rats and monkeys can be conducted using electrodes implanted in the brain, human subjects typically refuse to have their brain tissue exposed and to have wires implanted in their brains. The ideal human subjects for such studies are patients who have brain injuries or illnesses. In extreme cases of epilepsy which cannot be treated medicinally (about 4% of cases), the corpus callosum, which connects the hemispheres of the brain is "cut". In these patients, the brain is typically open for two weeks, and electrodes are implanted in the brain to monitor it. The patients are asked for permission to be studied during this period of time.
The two hemispheres of the brain control "handedness" in the body, with a hemisphere controlling the motor functions of the opposite side of the body. The brain is not completely symmetrical, though, for instance, the language center of the brain is typically in the left hemisphere. Dr Cerf showed a couple of videos in which a subject with a "split brain" had to sort tiles or cards with various symbols on them, with one hand at times "correcting" the sorting performed by the other hand.
Electrodes in the brain can detect the stimulation of a single neuron (Dr Cern cited the "Simpsons" neuron also mentioned by Dr
André Fenton in his SSC lecture). Here's a video of the "Simpsons neuron" in action:
A neuron that fires for a concept came to be known as a "Jennifer Aniston Cell"- one patient's neuron fired not only when an image of Jennifer Aniston was displayed, but also when the patient heard her name or saw it in written form. In another patient a similar cell fired when images, text, and spoken words represented the Sydney Opera House. This neuron was "fooled" when a display of the Baha'ai Lotus Temple was shown, but it stopped doing so when the discrepancies between images of the two buildings was pointed out to the subject.
In another experiment, a subject was asked to manipulate images on a monitor. The images chosen were of icon Marilyn Monroe and actor Josh Brolin- the subject was asked to concentrate on image, then the other, and the image on the monitor would shift accordingly. In essence, two neurons would vie for dominance. Wired has a great summary of the experiment with an accompanying video.
Dr Cerf then moved on to the role that neurons play in movement- in quadriplegics, the brain cells responsible for movement work properly, but patient is immobile. Is it possible to manipulate objects with the brain? Dr Cerf showed a very poignant video of a paralyzed woman who has been able to manipulate a robot arm with an electrode implanted in her motor cortex in order to drink her coffee- I challenge you to try to watch it without tearing up:
Currently, there are flaws in the work. Dr Cerf discussed, as Dr Anne Churchland did in her lecture, the fact that, although computers can "master" the game of chess, they have not been able to direct the physical act of moving the pieces on the board. The mechanics are being improved- in one particular experiment a monkey was able to manipulate a prosthetic arm via an electrode implanted in its brain in order to grab marshmallows:
During the trials, the monkey's arm was immobilized, necessitating the use of the prosthetic arm. Once able to manipulate the arm properly, the monkey continued to use the prosthetic even when both of its arms were free, effectively giving it a "third arm". Imagine the possibilities, an individual could use the TV remote, drink a beer, and scratch oneself all at once. Now THAT's what I call multitasking!
Of course, there are ethical problems as well as technical ones- one cannot simply implant electrodes into people without an underlying therapeutic need. The human subjects described by Dr Cerf were individuals who needed drastic medical interventions. That being said, it may be possible that the limitations of the human body may be overcome through the use of machines. He cited the transition of the typewriter, from Pellegrino Turri's machine to allow his blind lover to write letters to a ubiquitous office staple.
Dr Cerf then described an experiment in which individuals were presented with a "clock face" graphic, and were instructed to use a button to stop the clock "hand" whenever they chose, and were then asked to indicate when they felt the "urge" to stop the "clock". Using an EEG to measure brain activity, the test administrators determined that brain activity increased three seconds before the button was pressed, and 1.5 seconds before the subject indicated an "urge" to stop the clock. In subsequent iterations of the experiment, the subject was manipulated in different ways- they were told that they couldn't stop the "clock" under certain circumstances and were given conflicting stimuli. In some cases, the subjects tried to "fool" the administrators, but the EEG always gave the subject away.
Dr Cerf wrapped up by reiterating the whole topic of contradictory desires and behaviors. He cited the "snooze alarm" conundrum, in which one's desire to wake, characterized by the setting of the alarm, is "at war" with one's decision to sleep more. He showed various novelty alarm clocks which "escape" the snooze-prone. He wrapped up with a hilarious photo of fitness club patrons taking an escalator one flight to the gym. In the Q&A some bastard asked the good doctor if he'd seen any evidence that the increasing use of electronic gadgets was making subjects more comfortable with electrode implants. Dr Cerf indicated that, to the contrary, the trend is for more conservative therapeutic measures, with the use of electrode implants and radical brain surgery becoming less common as less drastic measures are put into place.
In the middle of the lecture, an individual at the other end of the room from myself fell ill, and there was a brief interruption of the talk, but Dr Cerf recovered quickly without losing his stride. Kudos to Dr Cerf for not missing a beat and delivering a superb lecture. It was another feather in the cap of the Secret Science Club. Here's a brief animated feature covering some of the topics of Dr Cerf's talk:
In other news, I had a conversation with the brilliant and awesome **FUTURE BLOG POST**, a previous Secret Science Club lecturer, about some assistance in an upcoming project. I am looking forward to this project, so watch this space. Additionally, Secret Science Goddess (one of a pantheon of two!) and chanteuse Dorian Devins will be conducting a dialog with Richard Dawkins on Wednesday, 9/25 at NYU's Skirball Center for the Performing Arts. A friend of mine who has just recently gotten into science via Dawkins' writings picked up tickets for a bunch of us to go, as soon as I texted him about the event. I feel bad about having told him about Dawkins' dickishness, but I feel that one needs to know about any warts that one's "heroes" may have. Me? I can see the beneficial accomplishments of Dr Dawkins as well as his dickery, and will be keenly interested in the possible controversies that will be brought to the fore in this forum. Of course, my primary reason for attending the event is my unwavering support for Dorian Devins, whose feminist credentials are as strong as her science-supporting community credentials. Put succinctly, I dig Devins more than Dawkins... nobody's ever had to make excuses for her behavior.
Dr Cerf's lecture dealt with decision making and the brain/body interface. Why do individuals want one thing, but do the opposite thing? At times, it seems that an individual is two people fighting for dominance in one body. How does the brain interact with the body in everyday life?
Dr Cerf illustrated this conundrum by relating the tragic tale of Charles Whitman, the infamous University of Texas "tower sniper". After the mass shooting, the police tried to determine a cause for the shooting spree. Whitman was generally seen as a quiet, law abiding man, but his diary told an alarming tale. Whitman wrote that he "feels like he's not the same guy" and described having violent urges. Before he embarked on his murderous foray, he indicated that he wanted an autopsy performed on him and left a check to pay for it. When an autopsy was performed, a tumor the size of a walnut was found in his brain which impinged on his amygdala.
I another instance, a patient with a frontal lobe tumor began to exhibit signs of sexual deviancy, including pedophila. The removal of the tumor caused him to revert back to his previous sexual mores, but a partial recurrence of the tumor caused a return of the symptoms until it too was removed.
To illustrate the ability of individuals to "color" their perceptions of events, Dr Cerf cited a study (PDF) in which which a group of colonoscopy patients were told to rate the pain they experienced on a ten-point pain scale. Half of the patients experienced an interval at the end of the procedure in which the tip of the scope remained in their rectums. Oddly enough, the patients with the extended colonoscopy rated their experience as less painful than the patients whose colonoscopies were shorter in duration. It is probable that they experienced fewer "highs and lows" during the procedure than those with a quicker colonoscopy, and that the patients with the quicker procedure assessed it from moment to moment, rather than assessing it as one single event.
Studying the "competing influences" in the brain poses some ethical dilemmas- while studies using rats and monkeys can be conducted using electrodes implanted in the brain, human subjects typically refuse to have their brain tissue exposed and to have wires implanted in their brains. The ideal human subjects for such studies are patients who have brain injuries or illnesses. In extreme cases of epilepsy which cannot be treated medicinally (about 4% of cases), the corpus callosum, which connects the hemispheres of the brain is "cut". In these patients, the brain is typically open for two weeks, and electrodes are implanted in the brain to monitor it. The patients are asked for permission to be studied during this period of time.
The two hemispheres of the brain control "handedness" in the body, with a hemisphere controlling the motor functions of the opposite side of the body. The brain is not completely symmetrical, though, for instance, the language center of the brain is typically in the left hemisphere. Dr Cerf showed a couple of videos in which a subject with a "split brain" had to sort tiles or cards with various symbols on them, with one hand at times "correcting" the sorting performed by the other hand.
Electrodes in the brain can detect the stimulation of a single neuron (Dr Cern cited the "Simpsons" neuron also mentioned by Dr
André Fenton in his SSC lecture). Here's a video of the "Simpsons neuron" in action:
A neuron that fires for a concept came to be known as a "Jennifer Aniston Cell"- one patient's neuron fired not only when an image of Jennifer Aniston was displayed, but also when the patient heard her name or saw it in written form. In another patient a similar cell fired when images, text, and spoken words represented the Sydney Opera House. This neuron was "fooled" when a display of the Baha'ai Lotus Temple was shown, but it stopped doing so when the discrepancies between images of the two buildings was pointed out to the subject.
In another experiment, a subject was asked to manipulate images on a monitor. The images chosen were of icon Marilyn Monroe and actor Josh Brolin- the subject was asked to concentrate on image, then the other, and the image on the monitor would shift accordingly. In essence, two neurons would vie for dominance. Wired has a great summary of the experiment with an accompanying video.
Dr Cerf then moved on to the role that neurons play in movement- in quadriplegics, the brain cells responsible for movement work properly, but patient is immobile. Is it possible to manipulate objects with the brain? Dr Cerf showed a very poignant video of a paralyzed woman who has been able to manipulate a robot arm with an electrode implanted in her motor cortex in order to drink her coffee- I challenge you to try to watch it without tearing up:
Currently, there are flaws in the work. Dr Cerf discussed, as Dr Anne Churchland did in her lecture, the fact that, although computers can "master" the game of chess, they have not been able to direct the physical act of moving the pieces on the board. The mechanics are being improved- in one particular experiment a monkey was able to manipulate a prosthetic arm via an electrode implanted in its brain in order to grab marshmallows:
During the trials, the monkey's arm was immobilized, necessitating the use of the prosthetic arm. Once able to manipulate the arm properly, the monkey continued to use the prosthetic even when both of its arms were free, effectively giving it a "third arm". Imagine the possibilities, an individual could use the TV remote, drink a beer, and scratch oneself all at once. Now THAT's what I call multitasking!
Of course, there are ethical problems as well as technical ones- one cannot simply implant electrodes into people without an underlying therapeutic need. The human subjects described by Dr Cerf were individuals who needed drastic medical interventions. That being said, it may be possible that the limitations of the human body may be overcome through the use of machines. He cited the transition of the typewriter, from Pellegrino Turri's machine to allow his blind lover to write letters to a ubiquitous office staple.
Dr Cerf then described an experiment in which individuals were presented with a "clock face" graphic, and were instructed to use a button to stop the clock "hand" whenever they chose, and were then asked to indicate when they felt the "urge" to stop the "clock". Using an EEG to measure brain activity, the test administrators determined that brain activity increased three seconds before the button was pressed, and 1.5 seconds before the subject indicated an "urge" to stop the clock. In subsequent iterations of the experiment, the subject was manipulated in different ways- they were told that they couldn't stop the "clock" under certain circumstances and were given conflicting stimuli. In some cases, the subjects tried to "fool" the administrators, but the EEG always gave the subject away.
Dr Cerf wrapped up by reiterating the whole topic of contradictory desires and behaviors. He cited the "snooze alarm" conundrum, in which one's desire to wake, characterized by the setting of the alarm, is "at war" with one's decision to sleep more. He showed various novelty alarm clocks which "escape" the snooze-prone. He wrapped up with a hilarious photo of fitness club patrons taking an escalator one flight to the gym. In the Q&A some bastard asked the good doctor if he'd seen any evidence that the increasing use of electronic gadgets was making subjects more comfortable with electrode implants. Dr Cerf indicated that, to the contrary, the trend is for more conservative therapeutic measures, with the use of electrode implants and radical brain surgery becoming less common as less drastic measures are put into place.
In the middle of the lecture, an individual at the other end of the room from myself fell ill, and there was a brief interruption of the talk, but Dr Cerf recovered quickly without losing his stride. Kudos to Dr Cerf for not missing a beat and delivering a superb lecture. It was another feather in the cap of the Secret Science Club. Here's a brief animated feature covering some of the topics of Dr Cerf's talk:
In other news, I had a conversation with the brilliant and awesome **FUTURE BLOG POST**, a previous Secret Science Club lecturer, about some assistance in an upcoming project. I am looking forward to this project, so watch this space. Additionally, Secret Science Goddess (one of a pantheon of two!) and chanteuse Dorian Devins will be conducting a dialog with Richard Dawkins on Wednesday, 9/25 at NYU's Skirball Center for the Performing Arts. A friend of mine who has just recently gotten into science via Dawkins' writings picked up tickets for a bunch of us to go, as soon as I texted him about the event. I feel bad about having told him about Dawkins' dickishness, but I feel that one needs to know about any warts that one's "heroes" may have. Me? I can see the beneficial accomplishments of Dr Dawkins as well as his dickery, and will be keenly interested in the possible controversies that will be brought to the fore in this forum. Of course, my primary reason for attending the event is my unwavering support for Dorian Devins, whose feminist credentials are as strong as her science-supporting community credentials. Put succinctly, I dig Devins more than Dawkins... nobody's ever had to make excuses for her behavior.
Subscribe to:
Posts (Atom)
