What do you need to build a self-driving car? Roboticists and computer scientists have generally settled on similar requirements. Your autonomous vehicle needs to know where the boundaries of the road are. It needs to be able to steer the car and hit the brakes. It needs to know the speed limit, be able to read street signs, and detect if a traffic light is red or green. It needs to be able to react quickly to unexpected objects in its path, and it gets extra points if it knows where it is on a map.
All of those skills are important and necessary. But by building from a list of technical requirements, researchers neglect the single most important part of real-world driving: our intuition. Using it to determine the motivations of those around us is something humans are so effortlessly good at that its hard to even notice were doing it, nonetheless program for it.
A self-driving car currently lacks the ability to look at a personwhether theyre walking, driving a car, or riding a bikeand know what theyre thinking. These instantaneous human judgments are vital to our safety when were drivingand to that of others on the road, too.
As the CTO and cofounder of Perceptive Automata, an autonomous-vehicle software company started by Harvard neuroscientists and computer scientists, I wanted to see how often humans make these kinds of subconscious calls on the road. I took a camera out to a calm intersection near my former lab at Harvard with no traffic signals. It is not by any stretch of the imagination as congested or difficult as an intersection in downtown Boston, let alone Manhattan or Mexico City. But in 30 seconds of video, it is still possible to count more than 45 instances of one person intuiting whats in the mind of another. These non-verbal, split-second intuitions could be that person is not going to yield, that person doesnt know Im here, or that person wouldnt jaywalk while walking a dog. Is that bicyclist going to turn left or stop? Is that pedestrian going to take advantage of their right-of-way and cross? These judgments happen instantaneously, just watch.
We have lots of empirical evidence that humans are incredibly good at intuiting the intentions of others. The Sally-Anne task is a classic psychology experiment. Subjectsusually childrenwatch a researcher acting out a scene with dolls. A doll named Sally hides a marble in a covered basket. Sally leaves the room. While Sally is gone, a second dollAnnesecretly moves the marble out of the basket and into a closed box. When the first doll comes back, children are asked where she will look for the marble. Its easy to say, Well, of course shell still look in the basket, as Sally couldnt have known that the marble had moved while she was gone. But that of course is hiding an immensely sophisticated model. Children have to know not only that Sally is aware of some things and not of others, but that her awareness only updates when she is able to pay attention to something. They also have to know that her mental state is persistent, even when she leaves the room and comes back. This task has been repeated many times in labs around the world, and is part of the standard toolkit researchers use to understand if somebodys social intuitions are intact.
The ability to predict the mental state of others is so innate that we even apply it to distinctly non-human objects. The Heider-Simel experiment shows how were prone to ascribe perceived intent even to simple geometric shapes. In this famous study, a film shows two triangles and a circle moving around the screen. With essentially no exceptions, most people construct and elaborate narrative about the goals and interactions of the geometric shapes: One is a villain, one a protector, the third a victim who grows courageous and saves the dayall these mental states and narratives just from looking at geometric shapes moving about. In the psychological literature, this is called an impoverished stimulus.
Our interactions with people using the road are an example of an impoverished stimulus, too. We only see a pedestrian for a few hundred milliseconds before we have to decide how to react to them. We see a car edging slightly into a lane for a half second and have to decide whether to yield to them. We catch a fleeting glimpse of a cyclist and judge whether they know were making a right turn. These kinds of interactions are constant, and they are at the very core of driving safely and considerately.
And computers, so far, are hopeless at navigating them.
The perils of lacking an intuition for state of mind are already evident. In the first at-fault crash of a self-driving vehicle, a Google self-driving car in Mountain View incorrectly assumed that a bus driver would yield to it, misunderstanding both the urgency and the flexibility of a human driver trying to get around a stopped vehicle. In another crash, a self-driving Uber in Arizona was hit by a turning driver who expected that any oncoming vehicles would notice the adjacent lanes of traffic had slowed down and adjust its expectations of how turning drivers would behave.
Why are computers so bad at this task of mind reading if its so easy for people? This circumstance comes up so often in AI development that it has a name: Moravecs Paradox. The tasks that are easiest for people are often the ones that are the hardest for computers. Were least aware of what our minds do best, said the late AI pioneer Marvin Minsky. Were more aware of simple processes that dont work well than of complex ones that work flawlessly.
So how do you design an algorithm to perform a task if you cant say with any certainty what the task entails?
The usual solution is to define the task as simply as possible and use what are called deep-learning algorithms that can learn from vast quantities of data. For example, when given a sufficient number of pictures of trees (and pictures of things that are not trees), these computer programs can do a very good job of identifying a tree. If you boil a problem down to either proving or disproving an unambiguous fact about the worldthere is a tree there, or there is notalgorithms can do a pretty good job.
The only way to solve these problems is to deeply understand human behavior by characterizing it carefully using the techniques of behavioral science.But what to do about problems where basic facts about the world are neither simple nor accessible? Humans can make surprisingly accurate judgments about other humans because we have an immensely sophisticated set of internal models for how those around us behave. But those models are hidden from scrutiny, hidden in the black boxes of our minds. How do you label images with the contents of somebodys constantly fluid and mostly nonsensical inner monologue?
The only way to solve these problems is to deeply understand human behaviornot just by reverse-engineering it, but by characterizing it carefully and comprehensively using the techniques of behavioral science. Humans are immensely capable but have opaque internal mechanisms. We need to use the techniques of human behavioral research in order to build computer-vision models that are trained to capture the nuances and subtleties of human responses to the world instead of trying to guess what our internal model of the world looks like.
First, we need to work out how humans worksecond comes training the machines. Only with a rich, deep characterization of the quirks and foibles of human ability can we know enough about the problem were trying to solve in order to build computer models that can solve it. By using humans as the model for ideal performance, we are able to gain traction on these difficult tasks and find a meaningful solution to this intuition problem.
And we need to solve it. If self-driving cars are going to achieve their promise as a revolution in urban transportationdelivering reduced emissions, better mobility, and safer streetsthey will have to exist on a level playing field with the humans who already use those roads. They will have to be good citize
ns, not only skilled at avoiding at-fault accidents, but able to drive in such a way that their behavior is expected, comprehensible, and clear to other vehicles drivers and the pedestrians and cyclists sharing space with them.
Follow Sam on Twitter. Learn how to write for Quartz Ideas. We welcome your comments at email@example.com.
Read more from the original source:
Self-driving cars still can't mimic the most natural human behavior - Quartz
- Seniors enter final year of school at UMFK - WAGM - September 8th, 2017
- Virtual Reality Can Make a Remote Crisis Real and Spur Effective Responses - University of Virginia - September 8th, 2017
- International research powerhouses join forces to advance study of life on Earth - Washington University in St. Louis Newsroom - September 6th, 2017
- Taking Advantage Of Behavioral Economics Can Get Aid To More ... - Fast Company - September 6th, 2017
- SRU completes busy summer of campus improvements - Allied News - September 4th, 2017
- Mayor Jim Kenney wants to clean up 'Filthadelphia' and he's enlisting Penn's help - The Daily Pennsylvanian - September 4th, 2017
- Meditation expert tells us what the science really says and why multitasking is a 'myth' - Southernminn.com - September 3rd, 2017
- Losing weight for the couch potato and others - The Washington Post - Washington Post - September 3rd, 2017
- Brain researchers in uproar over NIH clinical trials policy - Nature.com - September 3rd, 2017
- Mnoa volcanologists receive top international awards - UH System Current News - August 30th, 2017
- Developing technology with advisors at heart - Financial Planning - August 30th, 2017
- Text Messaging Initiative Will Nudge STEM Students Toward Success - Campus Technology - August 30th, 2017
- Sales Incentives and Machine Learning: Intelligently Motivate Revenue-Driving Behaviors - Customer Think - August 30th, 2017
- Basic studies of how our brains work are now clinical trials, NIH says - Science Magazine - August 27th, 2017
- How Overcoming Demands on Attention Can Help Alleviate Poverty - Newswise (press release) - August 24th, 2017
- Diverse programming and experiential learning top of mind for SEHHB Interim Dean Paul Rose - RiverBender.com - August 23rd, 2017
- Sylvia Sims Bolton appointed new Waukegan 1st Ward alderman - Chicago Tribune - August 23rd, 2017
- Science and Society on the Vineyard - Martha's Vineyard Times - August 23rd, 2017
- Scientists give star treatment to lesser-known cells crucial for brain development - Seacoastonline.com - August 20th, 2017
- Leveraging The Power of Behavioral Science in Banking - The Financial Brand - August 18th, 2017
- City Officials Turn To Behavioral Science To Improve Government Services - CBS Philly - August 18th, 2017
- Veteran's Disability Payments Compromised in Cyber Attack - NBC Chicago - August 11th, 2017
- MVC selects ISU's Goy for Hall of Fame - Bloomington Pantagraph - August 11th, 2017
- Suicide Rate Hit 40-Year Peak Among Older Teen Girls In 2015 - LEX18 Lexington KY News - August 6th, 2017
- The Case for Giving Health-Care Consumers a 'Nudge' - Wall Street Journal (subscription) - June 26th, 2017
- CPS draws on psychology to motivate customers to cut energy use in new program - mySanAntonio.com - June 12th, 2017
- Want To Employ Behavioral Science For Good? Here's A Helpful Collection Of Ideas - Fast Company - June 11th, 2017
- Ecologists protest sudden end of NSF dissertation grants - Science Magazine - June 9th, 2017
- Behavioral 'Nudges' Offer a Cost-Effective Policy Tool - Harvard Business School - June 9th, 2017
- New report: Social, behavioral, and economic sciences contribute to advancing NSF mission - Phys.Org - June 9th, 2017
- Top Schools for Behavioral Science - Study.com - June 8th, 2017
- Maritz Improves Business Performance with Innovative Behavioral Science - EQ - Entrepreneur Quarterly (press release) (subscription) (blog) - June 8th, 2017
- The Behavioral Economics of Why Executives Underinvest in Cybersecurity - Harvard Business Review - June 8th, 2017
- UL Lafayette to offer online General Studies degree program - KATC.com | Continuous News Coverage | Acadiana ... - KATC Lafayette News - June 8th, 2017
- Scientists use wearables to track patient data - Medical Xpress - June 8th, 2017
- Stanford Research on Sex Differences Reveals a Leftist Rejection of ... - Breitbart News - June 6th, 2017
- The New Way To Prevent Anxiety in Kids - TIME - June 6th, 2017
- Chapman University's Physical Therapy Program receives 10-year Accreditation - Chapman University: Happenings (blog) - June 6th, 2017
- Cops speak less respectfully to black community members - Stanford University News - June 6th, 2017
- What millennials really want in the workplace - CBS News - June 5th, 2017
- A Periodic Table of Behavior for Psychology - Psychology Today (blog) - June 4th, 2017
- LOOSE ENDS: Eldar Shafir on the effects scarcity - centraljersey.com - June 4th, 2017
- People trust science. So why don't they believe it? - WXIA-TV - June 4th, 2017
- Behavioral science hacks for your next speaking opportunity - SmartBrief (registration) (blog) - June 3rd, 2017
- SHOP TALK: Eldar Shafir on the effects scarcity - centraljersey.com - June 3rd, 2017
- People trust science. So why don't they believe it? - KING5.com - June 3rd, 2017
- Why Mainstream Media Need to Be Careful About Criticizing Conservatives - Patheos (blog) - June 3rd, 2017
- People trust science. So why don't they believe it? - WGRZ-TV - June 2nd, 2017
- UB program for underrepresented minority students in biomedical PhD programs wins coveted renewal - UB News Center - June 2nd, 2017
- Why Mainstream Media Need to Be Careful About Criticizing Conservatives - HuffPost - June 2nd, 2017
- Turn college debt into an investment - Green Bay Press Gazette - June 1st, 2017
- Are Behavioral Science, Customer Centricity And Customer Experience Compatible? - Forbes - June 1st, 2017
- To Counter Opioid Crisis, NIH Pushes Researchers to Invent More Drugs - The Chronicle of Higher Education (subscription) - June 1st, 2017
- Eco-Pass fee increase fight ramps up - La Voz Weekly - June 1st, 2017
- Scientists Want a Vaccine to Protect Readers From Fake News - Sputnik International - June 1st, 2017
- Behavioral neuroscience - Wikipedia - May 31st, 2017
- Wright State spinoff closes on $680K in funding - Dayton Business Journal - May 31st, 2017
- How To Navigate Your Child's Adolescence46:37 - WBUR - May 30th, 2017
- FRC Class of 2017 - Plumas County Newspapers - May 28th, 2017
- Can a Fidget Spinner Really Help You Focus? - Big Think - May 28th, 2017
- How Can Facts Trump Ideology? - Patheos (blog) - May 28th, 2017
- Nearly 1600 Santa Rosa Junior College students receive diplomas in class of 2017 - Santa Rosa Press Democrat - May 28th, 2017
- How Laws of Physics Govern Growth in Business and in Cities - New York Times - May 26th, 2017
- Lodi senior earns 9 associate degrees, picks UC Davis over 11 others - Sacramento Bee - May 26th, 2017
- Duke Hosts Precision Medicine World Conference - Duke Today (blog) - May 25th, 2017
- American Association of Behavioral and Social Sciences - May 25th, 2017
- Cognitive Behavioral Therapy| Psychiatry Conference ... - May 25th, 2017
- The president's budget proposal threatens science - The Hill (blog) - May 25th, 2017
- Celebrating Nearly 11,500 Graduates at CSUN Commencement ... - CSUN Today - May 25th, 2017
- Comments Off on Celebrating Nearly 11500 Graduates at CSUN Commencement 2017 - CSUN Today - May 24th, 2017
- Some Social Scientists Are Tired of Asking for Permission - New York Times - May 23rd, 2017
- Policymakers around the world are embracing behavioural science - The Economist - May 22nd, 2017
- Western Wayne students participate in PJAS competition - News ... - Scranton Times-Tribune - May 22nd, 2017
- What behavioral finance can teach us about markets and ourselves - InvestmentNews - May 22nd, 2017
- Behavioral Science - Psychology | Behavioral Science | Home - May 21st, 2017
- Sometimes, Facts Can Actually Trump Ideology - Scientific American (blog) - May 20th, 2017
- UVU's largest college appoints a new dean - Daily Herald - May 20th, 2017
- The effect of Moore's Law on behavioral marketing - MarTech Today - May 20th, 2017
- Charles Murray is once again peddling junk science about race and IQ - Vox - May 18th, 2017
- Fidget Toys Aren't Just Hype - Scientific American - May 18th, 2017