Tag Archives: researchers

#433776 Why We Should Stop Conflating Human and ...

Posted on October 19, 2018 by Android

It’s common to hear phrases like ‘machine learning’ and ‘artificial intelligence’ and believe that somehow, someone has managed to replicate a human mind inside a computer. This, of course, is untrue—but part of the reason this idea is so pervasive is because the metaphor of human learning and intelligence has been quite useful in explaining machine learning and artificial intelligence.

Indeed, some AI researchers maintain a close link with the neuroscience community, and inspiration runs in both directions. But the metaphor can be a hindrance to people trying to explain machine learning to those less familiar with it. One of the biggest risks of conflating human and machine intelligence is that we start to hand over too much agency to machines. For those of us working with software, it’s essential that we remember the agency is human—it’s humans who build these systems, after all.

It’s worth unpacking the key differences between machine and human intelligence. While there are certainly similarities, it’s by looking at what makes them different that we can better grasp how artificial intelligence works, and how we can build and use it effectively.

Neural Networks
Central to the metaphor that links human and machine learning is the concept of a neural network. The biggest difference between a human brain and an artificial neural net is the sheer scale of the brain’s neural network. What’s crucial is that it’s not simply the number of neurons in the brain (which reach into the billions), but more precisely, the mind-boggling number of connections between them.

But the issue runs deeper than questions of scale. The human brain is qualitatively different from an artificial neural network for two other important reasons: the connections that power it are analogue, not digital, and the neurons themselves aren’t uniform (as they are in an artificial neural network).

This is why the brain is such a complex thing. Even the most complex artificial neural network, while often difficult to interpret and unpack, has an underlying architecture and principles guiding it (this is what we’re trying to do, so let’s construct the network like this…).

Intricate as they may be, neural networks in AIs are engineered with a specific outcome in mind. The human mind, however, doesn’t have the same degree of intentionality in its engineering. Yes, it should help us do all the things we need to do to stay alive, but it also allows us to think critically and creatively in a way that doesn’t need to be programmed.

The Beautiful Simplicity of AI
The fact that artificial intelligence systems are so much simpler than the human brain is, ironically, what enables AIs to deal with far greater computational complexity than we can.

Artificial neural networks can hold much more information and data than the human brain, largely due to the type of data that is stored and processed in a neural network. It is discrete and specific, like an entry on an excel spreadsheet.

In the human brain, data doesn’t have this same discrete quality. So while an artificial neural network can process very specific data at an incredible scale, it isn’t able to process information in the rich and multidimensional manner a human brain can. This is the key difference between an engineered system and the human mind.

Despite years of research, the human mind still remains somewhat opaque. This is because the analog synaptic connections between neurons are almost impenetrable to the digital connections within an artificial neural network.

Speed and Scale
Consider what this means in practice. The relative simplicity of an AI allows it to do a very complex task very well, and very quickly. A human brain simply can’t process data at scale and speed in the way AIs need to if they’re, say, translating speech to text, or processing a huge set of oncology reports.

Essential to the way AI works in both these contexts is that it breaks data and information down into tiny constituent parts. For example, it could break sounds down into phonetic text, which could then be translated into full sentences, or break images into pieces to understand the rules of how a huge set of them is composed.

Humans often do a similar thing, and this is the point at which machine learning is most like human learning; like algorithms, humans break data or information into smaller chunks in order to process it.

But there’s a reason for this similarity. This breakdown process is engineered into every neural network by a human engineer. What’s more, the way this process is designed will be down to the problem at hand. How an artificial intelligence system breaks down a data set is its own way of ‘understanding’ it.

Even while running a highly complex algorithm unsupervised, the parameters of how an AI learns—how it breaks data down in order to process it—are always set from the start.

Human Intelligence: Defining Problems
Human intelligence doesn’t have this set of limitations, which is what makes us so much more effective at problem-solving. It’s the human ability to ‘create’ problems that makes us so good at solving them. There’s an element of contextual understanding and decision-making in the way humans approach problems.

AIs might be able to unpack problems or find new ways into them, but they can’t define the problem they’re trying to solve.

Algorithmic insensitivity has come into focus in recent years, with an increasing number of scandals around bias in AI systems. Of course, this is caused by the biases of those making the algorithms, but underlines the point that algorithmic biases can only be identified by human intelligence.

Human and Artificial Intelligence Should Complement Each Other
We must remember that artificial intelligence and machine learning aren’t simply things that ‘exist’ that we can no longer control. They are built, engineered, and designed by us. This mindset puts us in control of the future, and makes algorithms even more elegant and remarkable.

Image Credit: Liu zishan/Shutterstock Continue reading →

Posted in Human Robots

#433758 DeepMind’s New Research Plan to Make ...

Posted on October 15, 2018 by Android

Making sure artificial intelligence does what we want and behaves in predictable ways will be crucial as the technology becomes increasingly ubiquitous. It’s an area frequently neglected in the race to develop products, but DeepMind has now outlined its research agenda to tackle the problem.

AI safety, as the field is known, has been gaining prominence in recent years. That’s probably at least partly down to the overzealous warnings of a coming AI apocalypse from well-meaning, but underqualified pundits like Elon Musk and Stephen Hawking. But it’s also recognition of the fact that AI technology is quickly pervading all aspects of our lives, making decisions on everything from what movies we watch to whether we get a mortgage.

That’s why DeepMind hired a bevy of researchers who specialize in foreseeing the unforeseen consequences of the way we built AI back in 2016. And now the team has spelled out the three key domains they think require research if we’re going to build autonomous machines that do what we want.

In a new blog designed to provide updates on the team’s work, they introduce the ideas of specification, robustness, and assurance, which they say will act as the cornerstones of their future research. Specification involves making sure AI systems do what their operator intends; robustness means a system can cope with changes to its environment and attempts to throw it off course; and assurance involves our ability to understand what systems are doing and how to control them.

A classic thought experiment designed to illustrate how we could lose control of an AI system can help illustrate the problem of specification. Philosopher Nick Bostrom’s posited a hypothetical machine charged with making as many paperclips as possible. Because the creators fail to add what they might assume are obvious additional goals like not harming people, the AI wipes out humanity so we can’t switch it off before turning all matter in the universe into paperclips.

Obviously the example is extreme, but it shows how a poorly-specified goal can lead to unexpected and disastrous outcomes. Properly codifying the desires of the designer is no easy feat, though; often there are not neat ways to encompass both the explicit and implicit goals in ways that are understandable to the machine and don’t leave room for ambiguities, meaning we often rely on incomplete approximations.

The researchers note recent research by OpenAI in which an AI was trained to play a boat-racing game called CoastRunners. The game rewards players for hitting targets laid out along the race route. The AI worked out that it could get a higher score by repeatedly knocking over regenerating targets rather than actually completing the course. The blog post includes a link to a spreadsheet detailing scores of such examples.

Another key concern for AI designers is making their creation robust to the unpredictability of the real world. Despite their superhuman abilities on certain tasks, most cutting-edge AI systems are remarkably brittle. They tend to be trained on highly-curated datasets and so can fail when faced with unfamiliar input. This can happen by accident or by design—researchers have come up with numerous ways to trick image recognition algorithms into misclassifying things, including thinking a 3D printed tortoise was actually a gun.

Building systems that can deal with every possible encounter may not be feasible, so a big part of making AIs more robust may be getting them to avoid risks and ensuring they can recover from errors, or that they have failsafes to ensure errors don’t lead to catastrophic failure.

And finally, we need to have ways to make sure we can tell whether an AI is performing the way we expect it to. A key part of assurance is being able to effectively monitor systems and interpret what they’re doing—if we’re basing medical treatments or sentencing decisions on the output of an AI, we’d like to see the reasoning. That’s a major outstanding problem for popular deep learning approaches, which are largely indecipherable black boxes.

The other half of assurance is the ability to intervene if a machine isn’t behaving the way we’d like. But designing a reliable off switch is tough, because most learning systems have a strong incentive to prevent anyone from interfering with their goals.

The authors don’t pretend to have all the answers, but they hope the framework they’ve come up with can help guide others working on AI safety. While it may be some time before AI is truly in a position to do us harm, hopefully early efforts like these will mean it’s built on a solid foundation that ensures it is aligned with our goals.

Image Credit: cono0430 / Shutterstock.com Continue reading →

Posted in Human Robots

#433725 This Week’s Awesome Stories From ...

Posted on October 6, 2018 by Android

ROBOTICS
The Demise of Rethink Robotics Shows How Hard It Is to Make Machines Truly Smart
Will Knight | MIT Technology Review
“There’s growing interest in using recent advances in AI to make industrial robots a lot smarter and more useful. …But look carefully and you’ll see that these technologies are at a very early stage, and that deploying them commercially could prove extremely challenging. The demise of Rethink doesn’t mean industrial robotics isn’t flourishing, or that AI-driven advances won’t come about. But it shows just how hard doing real innovation in robotics can be.”

SCIENCE
The Human Cell Atlas Is Biologists’ Latest Grand Project
Megan Molteni | Wired
“Dubbed the Human Cell Atlas, the project intends to catalog all of the estimated 37 trillion cells that make up a human body. …By decoding the genes active in single cells, pegging different cell types to a specific address in the body, and tracing the molecular circuits between them, participating researchers plan to create a more comprehensive map of human biology than has ever existed before.”

TRANSPORTATION
US Will Rewrite Safety Rules to Permit Fully Driverless Cars on Public Roads
Andrew J. Hawkins | The Verge
“Under current US safety rules, a motor vehicle must have traditional controls, like a steering wheel, mirrors, and foot pedals, before it is allowed to operate on public roads. But that could all change under a new plan released on Thursday by the Department of Transportation that’s intended to open the floodgates for fully driverless cars.”

ARTIFICIAL INTELLIGENCE
When an AI Goes Full Jack Kerouac
Brian Merchant | The Atlantic
“By the end of the four-day trip, receipts emblazoned with artificially intelligent prose would cover the floor of the car. …it is a hallucinatory, oddly illuminating account of a bot’s life on the interstate; the Electric Kool-Aid Acid Test meets Google Street View, narrated by Siri.”

FUTURE OF FOOD
New Autonomous Farm Wants to Produce Food Without Human Workers
Erin Winick | MIT Technology Review
“As the firm’s cofounder Brandon Alexander puts it: ‘We are a farm and will always be a farm.’ But it’s no ordinary farm. For starters, the company’s 15 human employees share their work space with robots who quietly go about the business of tending rows and rows of leafy greens.”

Image Credit: Kotenko Olaksandr / Shutterstock.com Continue reading →

Posted in Human Robots

#433668 A Decade of Commercial Space ...

Posted on September 28, 2018 by Android

In many industries, a decade is barely enough time to cause dramatic change unless something disruptive comes along—a new technology, business model, or service design. The space industry has recently been enjoying all three.

But 10 years ago, none of those innovations were guaranteed. In fact, on Sept. 28, 2008, an entire company watched and hoped as their flagship product attempted a final launch after three failures. With cash running low, this was the last shot. Over 21,000 kilograms of kerosene and liquid oxygen ignited and powered two booster stages off the launchpad.

This first official picture of the Soviet satellite Sputnik I was issued in Moscow Oct. 9, 1957. The satellite measured 1 foot, 11 inches and weighed 184 pounds. The Space Age began as the Soviet Union launched Sputnik, the first man-made satellite, into orbit, on Oct. 4, 1957.AP Photo/TASS
When that Falcon 1 rocket successfully reached orbit and the company secured a subsequent contract with NASA, SpaceX had survived its ‘startup dip’. That milestone, the first privately developed liquid-fueled rocket to reach orbit, ignited a new space industry that is changing our world, on this planet and beyond. What has happened in the intervening years, and what does it mean going forward?

While scientists are busy developing new technologies that address the countless technical problems of space, there is another segment of researchers, including myself, studying the business angle and the operations issues facing this new industry. In a recent paper, my colleague Christopher Tang and I investigate the questions firms need to answer in order to create a sustainable space industry and make it possible for humans to establish extraterrestrial bases, mine asteroids and extend space travel—all while governments play an increasingly smaller role in funding space enterprises. We believe these business solutions may hold the less-glamorous key to unlocking the galaxy.

The New Global Space Industry
When the Soviet Union launched their Sputnik program, putting a satellite in orbit in 1957, they kicked off a race to space fueled by international competition and Cold War fears. The Soviet Union and the United States played the primary roles, stringing together a series of “firsts” for the record books. The first chapter of the space race culminated with Neil Armstrong and Buzz Aldrin’s historic Apollo 11 moon landing which required massive public investment, on the order of US$25.4 billion, almost $200 billion in today’s dollars.

Competition characterized this early portion of space history. Eventually, that evolved into collaboration, with the International Space Station being a stellar example, as governments worked toward shared goals. Now, we’ve entered a new phase—openness—with private, commercial companies leading the way.

The industry for spacecraft and satellite launches is becoming more commercialized, due, in part, to shrinking government budgets. According to a report from the investment firm Space Angels, a record 120 venture capital firms invested over $3.9 billion in private space enterprises last year. The space industry is also becoming global, no longer dominated by the Cold War rivals, the United States and USSR.

In 2018 to date, there have been 72 orbital launches, an average of two per week, from launch pads in China, Russia, India, Japan, French Guinea, New Zealand, and the US.

The uptick in orbital launches of actual rockets as well as spacecraft launches, which includes satellites and probes launched from space, coincides with this openness over the past decade.

More governments, firms and even amateurs engage in various spacecraft launches than ever before. With more entities involved, innovation has flourished. As Roberson notes in Digital Trends, “Private, commercial spaceflight. Even lunar exploration, mining, and colonization—it’s suddenly all on the table, making the race for space today more vital than it has felt in years.”

Worldwide launches into space. Orbital launches include manned and unmanned spaceships launched into orbital flight from Earth. Spacecraft launches include all vehicles such as spaceships, satellites and probes launched from Earth or space. Wooten, J. and C. Tang (2018) Operations in space, Decision Sciences; Space Launch Report (Kyle 2017); Spacecraft Encyclopedia (Lafleur 2017), CC BY-ND

One can see this vitality plainly in the news. On Sept. 21, Japan announced that two of its unmanned rovers, dubbed Minerva-II-1, had landed on a small, distant asteroid. For perspective, the scale of this landing is similar to hitting a 6-centimeter target from 20,000 kilometers away. And earlier this year, people around the world watched in awe as SpaceX’s Falcon Heavy rocket successfully launched and, more impressively, returned its two boosters to a landing pad in a synchronized ballet of epic proportions.

Challenges and Opportunities
Amidst the growth of capital, firms, and knowledge, both researchers and practitioners must figure out how entities should manage their daily operations, organize their supply chain, and develop sustainable operations in space. This is complicated by the hurdles space poses: distance, gravity, inhospitable environments, and information scarcity.

One of the greatest challenges involves actually getting the things people want in space, into space. Manufacturing everything on Earth and then launching it with rockets is expensive and restrictive. A company called Made In Space is taking a different approach by maintaining an additive manufacturing facility on the International Space Station and 3D printing right in space. Tools, spare parts, and medical devices for the crew can all be created on demand. The benefits include more flexibility and better inventory management on the space station. In addition, certain products can be produced better in space than on Earth, such as pure optical fiber.

How should companies determine the value of manufacturing in space? Where should capacity be built and how should it be scaled up? The figure below breaks up the origin and destination of goods between Earth and space and arranges products into quadrants. Humans have mastered the lower left quadrant, made on Earth—for use on Earth. Moving clockwise from there, each quadrant introduces new challenges, for which we have less and less expertise.

A framework of Earth-space operations. Wooten, J. and C. Tang (2018) Operations in Space, Decision Sciences, CC BY-ND
I first became interested in this particular problem as I listened to a panel of robotics experts discuss building a colony on Mars (in our third quadrant). You can’t build the structures on Earth and easily send them to Mars, so you must manufacture there. But putting human builders in that extreme environment is equally problematic. Essentially, an entirely new mode of production using robots and automation in an advance envoy may be required.

Resources in Space
You might wonder where one gets the materials for manufacturing in space, but there is actually an abundance of resources: Metals for manufacturing can be found within asteroids, water for rocket fuel is frozen as ice on planets and moons, and rare elements like helium-3 for energy are embedded in the crust of the moon. If we brought that particular isotope back to Earth, we could eliminate our dependence on fossil fuels.

As demonstrated by the recent Minerva-II-1 asteroid landing, people are acquiring the technical know-how to locate and navigate to these materials. But extraction and transport are open questions.

How do these cases change the economics in the space industry? Already, companies like Planetary Resources, Moon Express, Deep Space Industries, and Asterank are organizing to address these opportunities. And scholars are beginning to outline how to navigate questions of property rights, exploitation and partnerships.

Threats From Space Junk
A computer-generated image of objects in Earth orbit that are currently being tracked. Approximately 95 percent of the objects in this illustration are orbital debris – not functional satellites. The dots represent the current location of each item. The orbital debris dots are scaled according to the image size of the graphic to optimize their visibility and are not scaled to Earth. NASA
The movie “Gravity” opens with a Russian satellite exploding, which sets off a chain reaction of destruction thanks to debris hitting a space shuttle, the Hubble telescope, and part of the International Space Station. The sequence, while not perfectly plausible as written, is a very real phenomenon. In fact, in 2013, a Russian satellite disintegrated when it was hit with fragments from a Chinese satellite that exploded in 2007. Known as the Kessler effect, the danger from the 500,000-plus pieces of space debris has already gotten some attention in public policy circles. How should one prevent, reduce or mitigate this risk? Quantifying the environmental impact of the space industry and addressing sustainable operations is still to come.

NASA scientist Mark Matney is seen through a fist-sized hole in a 3-inch thick piece of aluminum at Johnson Space Center’s orbital debris program lab. The hole was created by a thumb-size piece of material hitting the metal at very high speed simulating possible damage from space junk. AP Photo/Pat Sullivan
What’s Next?
It’s true that space is becoming just another place to do business. There are companies that will handle the logistics of getting your destined-for-space module on board a rocket; there are companies that will fly those rockets to the International Space Station; and there are others that can make a replacement part once there.

What comes next? In one sense, it’s anybody’s guess, but all signs point to this new industry forging ahead. A new breakthrough could alter the speed, but the course seems set: exploring farther away from home, whether that’s the moon, asteroids, or Mars. It’s hard to believe that 10 years ago, SpaceX launches were yet to be successful. Today, a vibrant private sector consists of scores of companies working on everything from commercial spacecraft and rocket propulsion to space mining and food production. The next step is working to solidify the business practices and mature the industry.

Standing in a large hall at the University of Pittsburgh as part of the White House Frontiers Conference, I see the future. Wrapped around my head are state-of-the-art virtual reality goggles. I’m looking at the surface of Mars. Every detail is immediate and crisp. This is not just a video game or an aimless exercise. The scientific community has poured resources into such efforts because exploration is preceded by information. And who knows, maybe 10 years from now, someone will be standing on the actual surface of Mars.

Image Credit: SpaceX

Joel Wooten, Assistant Professor of Management Science, University of South Carolina

This article is republished from The Conversation under a Creative Commons license. Read the original article. Continue reading →

Posted in Human Robots

#433486 This AI Predicts Obesity ...

Posted on September 17, 2018 by Android

A research team at the University of Washington has trained an artificial intelligence system to spot obesity—all the way from space. The system used a convolutional neural network (CNN) to analyze 150,000 satellite images and look for correlations between the physical makeup of a neighborhood and the prevalence of obesity.

The team’s results, presented in JAMA Network Open, showed that features of a given neighborhood could explain close to two-thirds (64.8 percent) of the variance in obesity. Researchers found that analyzing satellite data could help increase understanding of the link between peoples’ environment and obesity prevalence. The next step would be to make corresponding structural changes in the way neighborhoods are built to encourage physical activity and better health.

Training AI to Spot Obesity
Convolutional neural networks (CNNs) are particularly adept at image analysis, object recognition, and identifying special hierarchies in large datasets.

Prior to analyzing 150,000 high-resolution satellite images of Bellevue, Seattle, Tacoma, Los Angeles, Memphis, and San Antonio, the researchers trained the CNN on 1.2 million images from the ImageNet database. The categorizations were correlated with obesity prevalence estimates for the six urban areas from census tracts gathered by the 500 Cities project.

The system was able to identify the presence of certain features that increased likelihood of obesity in a given area. Some of these features included tightly–packed houses, being close to roadways, and living in neighborhoods with a lack of greenery.

Visualization of features identified by the convolutional neural network (CNN) model. The images on the left column are satellite images taken from Google Static Maps API (application programming interface). Images in the middle and right columns are activation maps taken from the second convolutional layer of VGG-CNN-F network after forward pass of the respective satellite images through the network. From Google Static Maps API, DigitalGlobe, US Geological Survey (accessed July 2017). Credit: JAMA Network Open
Your Surroundings Are Key
In their discussion of the findings, the researchers stressed that there are limitations to the conclusions that can be drawn from the AI’s results. For example, socio-economic factors like income likely play a major role for obesity prevalence in a given geographic area.

However, the study concluded that the AI-powered analysis showed the prevalence of specific man-made features in neighborhoods consistently correlating with obesity prevalence and not necessarily correlating with socioeconomic status.

The system’s success rates varied between studied cities, with Memphis being the highest (73.3 percent) and Seattle being the lowest (55.8 percent).

AI Takes To the Sky
Around a third of the US population is categorized as obese. Obesity is linked to a number of health-related issues, and the AI-generated results could potentially help improve city planning and better target campaigns to limit obesity.

The study is one of the latest of a growing list that uses AI to analyze images and extrapolate insights.

A team at Stanford University has used a CNN to predict poverty via satellite imagery, assisting governments and NGOs to better target their efforts. A combination of the public Automatic Identification System for shipping, satellite imagery, and Google’s AI has proven able to identify illegal fishing activity. Researchers have even been able to use AI and Google Street View to predict what party a given city will vote for, based on what cars are parked on the streets.

In each case, the AI systems have been able to look at volumes of data about our world and surroundings that are beyond the capabilities of humans and extrapolate new insights. If one were to moralize about the good and bad sides of AI (new opportunities vs. potential job losses, for example) it could seem that it comes down to what we ask AI systems to look at—and what questions we ask of them.

Image Credit: Ocean Biology Processing Group at NASA’s Goddard Space Flight Center Continue reading →

Posted in Human Robots