Tag Archives: operating
#434151 Life-or-Death Algorithms: The Black Box ...
When it comes to applications for machine learning, few can be more widely hyped than medicine. This is hardly surprising: it’s a huge industry that generates a phenomenal amount of data and revenue, where technological advances can improve or save the lives of millions of people. Hardly a week passes without a study that suggests algorithms will soon be better than experts at detecting pneumonia, or Alzheimer’s—diseases in complex organs ranging from the eye to the heart.
The problems of overcrowded hospitals and overworked medical staff plague public healthcare systems like Britain’s NHS and lead to rising costs for private healthcare systems. Here, again, algorithms offer a tantalizing solution. How many of those doctor’s visits really need to happen? How many could be replaced by an interaction with an intelligent chatbot—especially if it can be combined with portable diagnostic tests, utilizing the latest in biotechnology? That way, unnecessary visits could be reduced, and patients could be diagnosed and referred to specialists more quickly without waiting for an initial consultation.
As ever with artificial intelligence algorithms, the aim is not to replace doctors, but to give them tools to reduce the mundane or repetitive parts of the job. With an AI that can examine thousands of scans in a minute, the “dull drudgery” is left to machines, and the doctors are freed to concentrate on the parts of the job that require more complex, subtle, experience-based judgement of the best treatments and the needs of the patient.
High Stakes
But, as ever with AI algorithms, there are risks involved with relying on them—even for tasks that are considered mundane. The problems of black-box algorithms that make inexplicable decisions are bad enough when you’re trying to understand why that automated hiring chatbot was unimpressed by your job interview performance. In a healthcare context, where the decisions made could mean life or death, the consequences of algorithmic failure could be grave.
A new paper in Science Translational Medicine, by Nicholson Price, explores some of the promises and pitfalls of using these algorithms in the data-rich medical environment.
Neural networks excel at churning through vast quantities of training data and making connections, absorbing the underlying patterns or logic for the system in hidden layers of linear algebra; whether it’s detecting skin cancer from photographs or learning to write in pseudo-Shakespearean script. They are terrible, however, at explaining the underlying logic behind the relationships that they’ve found: there is often little more than a string of numbers, the statistical “weights” between the layers. They struggle to distinguish between correlation and causation.
This raises interesting dilemmas for healthcare providers. The dream of big data in medicine is to feed a neural network on “huge troves of health data, finding complex, implicit relationships and making individualized assessments for patients.” What if, inevitably, such an algorithm proves to be unreasonably effective at diagnosing a medical condition or prescribing a treatment, but you have no scientific understanding of how this link actually works?
Too Many Threads to Unravel?
The statistical models that underlie such neural networks often assume that variables are independent of each other, but in a complex, interacting system like the human body, this is not always the case.
In some ways, this is a familiar concept in medical science—there are many phenomena and links which have been observed for decades but are still poorly understood on a biological level. Paracetamol is one of the most commonly-prescribed painkillers, but there’s still robust debate about how it actually works. Medical practitioners may be keen to deploy whatever tool is most effective, regardless of whether it’s based on a deeper scientific understanding. Fans of the Copenhagen interpretation of quantum mechanics might spin this as “Shut up and medicate!”
But as in that field, there’s a debate to be had about whether this approach risks losing sight of a deeper understanding that will ultimately prove more fruitful—for example, for drug discovery.
Away from the philosophical weeds, there are more practical problems: if you don’t understand how a black-box medical algorithm is operating, how should you approach the issues of clinical trials and regulation?
Price points out that, in the US, the “21st-Century Cures Act” allows the FDA to regulate any algorithm that analyzes images, or doesn’t allow a provider to review the basis for its conclusions: this could completely exclude “black-box” algorithms of the kind described above from use.
Transparency about how the algorithm functions—the data it looks at, and the thresholds for drawing conclusions or providing medical advice—may be required, but could also conflict with the profit motive and the desire for secrecy in healthcare startups.
One solution might be to screen algorithms that can’t explain themselves, or don’t rely on well-understood medical science, from use before they enter the healthcare market. But this could prevent people from reaping the benefits that they can provide.
Evaluating Algorithms
New healthcare algorithms will be unable to do what physicists did with quantum mechanics, and point to a track record of success, because they will not have been deployed in the field. And, as Price notes, many algorithms will improve as they’re deployed in the field for a greater amount of time, and can harvest and learn from the performance data that’s actually used. So how can we choose between the most promising approaches?
Creating a standardized clinical trial and validation system that’s equally valid across algorithms that function in different ways, or use different input or training data, will be a difficult task. Clinical trials that rely on small sample sizes, such as for algorithms that attempt to personalize treatment to individuals, will also prove difficult. With a small sample size and little scientific understanding, it’s hard to tell whether the algorithm succeeded or failed because it’s bad at its job or by chance.
Add learning into the mix and the picture gets more complex. “Perhaps more importantly, to the extent that an ideal black-box algorithm is plastic and frequently updated, the clinical trial validation model breaks down further, because the model depends on a static product subject to stable validation.” As Price describes, the current system for testing and validation of medical products needs some adaptation to deal with this new software before it can successfully test and validate the new algorithms.
Striking a Balance
The story in healthcare reflects the AI story in so many other fields, and the complexities involved perhaps illustrate why even an illustrious company like IBM appears to be struggling to turn its famed Watson AI into a viable product in the healthcare space.
A balance must be struck, both in our rush to exploit big data and the eerie power of neural networks, and to automate thinking. We must be aware of the biases and flaws of this approach to problem-solving: to realize that it is not a foolproof panacea.
But we also need to embrace these technologies where they can be a useful complement to the skills, insights, and deeper understanding that humans can provide. Much like a neural network, our industries need to train themselves to enhance this cooperation in the future.
Image Credit: Connect world / Shutterstock.com Continue reading →
Posted in Human Robots
#433954 The Next Great Leap Forward? Combining ...
The Internet of Things is a popular vision of objects with internet connections sending information back and forth to make our lives easier and more comfortable. It’s emerging in our homes, through everything from voice-controlled speakers to smart temperature sensors. To improve our fitness, smart watches and Fitbits are telling online apps how much we’re moving around. And across entire cities, interconnected devices are doing everything from increasing the efficiency of transport to flood detection.
In parallel, robots are steadily moving outside the confines of factory lines. They’re starting to appear as guides in shopping malls and cruise ships, for instance. As prices fall and the artificial intelligence (AI) and mechanical technology continues to improve, we will get more and more used to them making independent decisions in our homes, streets and workplaces.
Here lies a major opportunity. Robots become considerably more capable with internet connections. There is a growing view that the next evolution of the Internet of Things will be to incorporate them into the network, opening up thrilling possibilities along the way.
Home Improvements
Even simple robots become useful when connected to the internet—getting updates about their environment from sensors, say, or learning about their users’ whereabouts and the status of appliances in the vicinity. This lets them lend their bodies, eyes, and ears to give an otherwise impersonal smart environment a user-friendly persona. This can be particularly helpful for people at home who are older or have disabilities.
We recently unveiled a futuristic apartment at Heriot-Watt University to work on such possibilities. One of a few such test sites around the EU, our whole focus is around people with special needs—and how robots can help them by interacting with connected devices in a smart home.
Suppose a doorbell rings that has smart video features. A robot could find the person in the home by accessing their location via sensors, then tell them who is at the door and why. Or it could help make video calls to family members or a professional carer—including allowing them to make virtual visits by acting as a telepresence platform.
Equally, it could offer protection. It could inform them the oven has been left on, for example—phones or tablets are less reliable for such tasks because they can be misplaced or not heard.
Similarly, the robot could raise the alarm if its user appears to be in difficulty.Of course, voice-assistant devices like Alexa or Google Home can offer some of the same services. But robots are far better at moving, sensing and interacting with their environment. They can also engage their users by pointing at objects or acting more naturally, using gestures or facial expressions. These “social abilities” create bonds which are crucially important for making users more accepting of the support and making it more effective.
To help incentivize the various EU test sites, our apartment also hosts the likes of the European Robotic League Service Robot Competition—a sort of Champions League for robots geared to special needs in the home. This brought academics from around Europe to our laboratory for the first time in January this year. Their robots were tested in tasks like welcoming visitors to the home, turning the oven off, and fetching objects for their users; and a German team from Koblenz University won with a robot called Lisa.
Robots Offshore
There are comparable opportunities in the business world. Oil and gas companies are looking at the Internet of Things, for example; experimenting with wireless sensors to collect information such as temperature, pressure, and corrosion levels to detect and possibly predict faults in their offshore equipment.
In the future, robots could be alerted to problem areas by sensors to go and check the integrity of pipes and wells, and to make sure they are operating as efficiently and safely as possible. Or they could place sensors in parts of offshore equipment that are hard to reach, or help to calibrate them or replace their batteries.
The likes of the ORCA Hub, a £36m project led by the Edinburgh Centre for Robotics, bringing together leading experts and over 30 industry partners, is developing such systems. The aim is to reduce the costs and the risks of humans working in remote hazardous locations.
ORCA tests a drone robot. ORCA
Working underwater is particularly challenging, since radio waves don’t move well under the sea. Underwater autonomous vehicles and sensors usually communicate using acoustic waves, which are many times slower (1,500 meters a second vs. 300m meters a second for radio waves). Acoustic communication devices are also much more expensive than those used above the water.
This academic project is developing a new generation of low-cost acoustic communication devices, and trying to make underwater sensor networks more efficient. It should help sensors and underwater autonomous vehicles to do more together in future—repair and maintenance work similar to what is already possible above the water, plus other benefits such as helping vehicles to communicate with one another over longer distances and tracking their location.
Beyond oil and gas, there is similar potential in sector after sector. There are equivalents in nuclear power, for instance, and in cleaning and maintaining the likes of bridges and buildings. My colleagues and I are also looking at possibilities in areas such as farming, manufacturing, logistics, and waste.
First, however, the research sectors around the Internet of Things and robotics need to properly share their knowledge and expertise. They are often isolated from one another in different academic fields. There needs to be more effort to create a joint community, such as the dedicated workshops for such collaboration that we organized at the European Robotics Forum and the IoT Week in 2017.
To the same end, industry and universities need to look at setting up joint research projects. It is particularly important to address safety and security issues—hackers taking control of a robot and using it to spy or cause damage, for example. Such issues could make customers wary and ruin a market opportunity.
We also need systems that can work together, rather than in isolated applications. That way, new and more useful services can be quickly and effectively introduced with no disruption to existing ones. If we can solve such problems and unite robotics and the Internet of Things, it genuinely has the potential to change the world.
Mauro Dragone, Assistant Professor, Cognitive Robotics, Multiagent systems, Internet of Things, Heriot-Watt University
This article is republished from The Conversation under a Creative Commons license. Read the original article.
Image Credit: Willyam Bradberry/Shutterstock.com Continue reading →
Posted in Human Robots
#433895 Sci-Fi Movies Are the Secret Weapon That ...
If there’s one line that stands the test of time in Steven Spielberg’s 1993 classic Jurassic Park, it’s probably Jeff Goldblum’s exclamation, “Your scientists were so preoccupied with whether or not they could, they didn’t stop to think if they should.”
Goldblum’s character, Dr. Ian Malcolm, was warning against the hubris of naively tinkering with dinosaur DNA in an effort to bring these extinct creatures back to life. Twenty-five years on, his words are taking on new relevance as a growing number of scientists and companies are grappling with how to tread the line between “could” and “should” in areas ranging from gene editing and real-world “de-extinction” to human augmentation, artificial intelligence and many others.
Despite growing concerns that powerful emerging technologies could lead to unexpected and wide-ranging consequences, innovators are struggling with how to develop beneficial new products while being socially responsible. Part of the answer could lie in watching more science fiction movies like Jurassic Park.
Hollywood Lessons in Societal Risks
I’ve long been interested in how innovators and others can better understand the increasingly complex landscape around the social risks and benefits associated with emerging technologies. Growing concerns over the impacts of tech on jobs, privacy, security and even the ability of people to live their lives without undue interference highlight the need for new thinking around how to innovate responsibly.
New ideas require creativity and imagination, and a willingness to see the world differently. And this is where science fiction movies can help.
Sci-fi flicks are, of course, notoriously unreliable when it comes to accurately depicting science and technology. But because their plots are often driven by the intertwined relationships between people and technology, they can be remarkably insightful in revealing social factors that affect successful and responsible innovation.
This is clearly seen in Jurassic Park. The movie provides a surprisingly good starting point for thinking about the pros and cons of modern-day genetic engineering and the growing interest in bringing extinct species back from the dead. But it also opens up conversations around the nature of complex systems that involve both people and technology, and the potential dangers of “permissionless” innovation that’s driven by power, wealth and a lack of accountability.
Similar insights emerge from a number of other movies, including Spielberg’s 2002 film “Minority Report”—which presaged a growing capacity for AI-enabled crime prediction and the ethical conundrums it’s raising—as well as the 2014 film Ex Machina.
As with Jurassic Park, Ex Machina centers around a wealthy and unaccountable entrepreneur who is supremely confident in his own abilities. In this case, the technology in question is artificial intelligence.
The movie tells a tale of an egotistical genius who creates a remarkable intelligent machine—but he lacks the awareness to recognize his limitations and the risks of what he’s doing. It also provides a chilling insight into potential dangers of creating machines that know us better than we know ourselves, while not being bound by human norms or values.
The result is a sobering reminder of how, without humility and a good dose of humanity, our innovations can come back to bite us.
The technologies in Jurassic Park, Minority Report, and Ex Machina lie beyond what is currently possible. Yet these films are often close enough to emerging trends that they help reveal the dangers of irresponsible, or simply naive, innovation. This is where these and other science fiction movies can help innovators better understand the social challenges they face and how to navigate them.
Real-World Problems Worked Out On-Screen
In a recent op-ed in the New York Times, journalist Kara Swisher asked, “Who will teach Silicon Valley to be ethical?” Prompted by a growing litany of socially questionable decisions amongst tech companies, Swisher suggests that many of them need to grow up and get serious about ethics. But ethics alone are rarely enough. It’s easy for good intentions to get swamped by fiscal pressures and mired in social realities.
Elon Musk has shown that brilliant tech innovators can take ethical missteps along the way. Image Credit:AP Photo/Chris Carlson
Technology companies increasingly need to find some way to break from business as usual if they are to become more responsible. High-profile cases involving companies like Facebook and Uber as well as Tesla’s Elon Musk have highlighted the social as well as the business dangers of operating without fully understanding the consequences of people-oriented actions.
Many more companies are struggling to create socially beneficial technologies and discovering that, without the necessary insights and tools, they risk blundering about in the dark.
For instance, earlier this year, researchers from Google and DeepMind published details of an artificial intelligence-enabled system that can lip-read far better than people. According to the paper’s authors, the technology has enormous potential to improve the lives of people who have trouble speaking aloud. Yet it doesn’t take much to imagine how this same technology could threaten the privacy and security of millions—especially when coupled with long-range surveillance cameras.
Developing technologies like this in socially responsible ways requires more than good intentions or simply establishing an ethics board. People need a sophisticated understanding of the often complex dynamic between technology and society. And while, as Mozilla’s Mitchell Baker suggests, scientists and technologists engaging with the humanities can be helpful, it’s not enough.
An Easy Way into a Serious Discipline
The “new formulation” of complementary skills Baker says innovators desperately need already exists in a thriving interdisciplinary community focused on socially responsible innovation. My home institution, the School for the Future of Innovation in Society at Arizona State University, is just one part of this.
Experts within this global community are actively exploring ways to translate good ideas into responsible practices. And this includes the need for creative insights into the social landscape around technology innovation, and the imagination to develop novel ways to navigate it.
People love to come together as a movie audience.Image credit: The National Archives UK, CC BY 4.0
Here is where science fiction movies become a powerful tool for guiding innovators, technology leaders and the companies where they work. Their fictional scenarios can reveal potential pitfalls and opportunities that can help steer real-world decisions toward socially beneficial and responsible outcomes, while avoiding unnecessary risks.
And science fiction movies bring people together. By their very nature, these films are social and educational levelers. Look at who’s watching and discussing the latest sci-fi blockbuster, and you’ll often find a diverse cross-section of society. The genre can help build bridges between people who know how science and technology work, and those who know what’s needed to ensure they work for the good of society.
This is the underlying theme in my new book Films from the Future: The Technology and Morality of Sci-Fi Movies. It’s written for anyone who’s curious about emerging trends in technology innovation and how they might potentially affect society. But it’s also written for innovators who want to do the right thing and just don’t know where to start.
Of course, science fiction films alone aren’t enough to ensure socially responsible innovation. But they can help reveal some profound societal challenges facing technology innovators and possible ways to navigate them. And what better way to learn how to innovate responsibly than to invite some friends round, open the popcorn and put on a movie?
It certainly beats being blindsided by risks that, with hindsight, could have been avoided.
Andrew Maynard, Director, Risk Innovation Lab, Arizona State University
This article is republished from The Conversation under a Creative Commons license. Read the original article.
Image Credit: Fred Mantel / Shutterstock.com Continue reading →
Posted in Human Robots
#433698 Microsoft Announces Experimental Release ...
Microsoft takes a step back into robotics by bringing the Robot Operating System to Windows Continue reading →
Posted in Human Robots
#433655 First-Ever Grad Program in Space Mining ...
Maybe they could call it the School of Space Rock: A new program being offered at the Colorado School of Mines (CSM) will educate post-graduate students on the nuts and bolts of extracting and using valuable materials such as rare metals and frozen water from space rocks like asteroids or the moon.
Officially called Space Resources, the graduate-level program is reputedly the first of its kind in the world to offer a course in the emerging field of space mining. Heading the program is Angel Abbud-Madrid, director of the Center for Space Resources at Mines, a well-known engineering school located in Golden, Colorado, where Molson Coors taps Rocky Mountain spring water for its earthly brews.
The first semester for the new discipline began last month. While Abbud-Madrid didn’t immediately respond to an interview request, Singularity Hub did talk to Chris Lewicki, president and CEO of Planetary Resources, a space mining company whose founders include Peter Diamandis, Singularity University co-founder.
A former NASA engineer who worked on multiple Mars missions, Lewicki says the Space Resources program at CSM, with its multidisciplinary focus on science, economics, and policy, will help students be light years ahead of their peers in the nascent field of space mining.
“I think it’s very significant that they’ve started this program,” he said. “Having students with that kind of background exposure just allows them to be productive on day one instead of having to kind of fill in a lot of things for them.”
Who would be attracted to apply for such a program? There are many professionals who could be served by a post-baccalaureate certificate, master’s degree, or even Ph.D. in Space Resources, according to Lewicki. Certainly aerospace engineers and planetary scientists would be among the faces in the classroom.
“I think it’s [also] people who have an interest in what I would call maybe space robotics,” he said. Lewicki is referring not only to the classic example of robotic arms like the Canadarm2, which lends a hand to astronauts aboard the International Space Station, but other types of autonomous platforms.
One example might be Planetary Resources’ own Arkyd-6, a small, autonomous satellite called a CubeSat launched earlier this year to test different technologies that might be used for deep-space exploration of resources. The proof-of-concept was as much a test for the technology—such as the first space-based use of a mid-wave infrared imager to detect water resources—as it was for being able to work in space on a shoestring budget.
“We really proved that doing one of these billion-dollar science missions to deep space can be done for a lot less if you have a very focused goal, and if you kind of cut a lot of corners and then put some commercial approaches into those things,” Lewicki said.
A Trillion-Dollar Industry
Why space mining? There are at least a trillion reasons.
Astrophysicist Neil deGrasse Tyson famously said that the first trillionaire will be the “person who exploits the natural resources on asteroids.” That’s because asteroids—rocky remnants from the formation of our solar system more than four billion years ago—harbor precious metals, ranging from platinum and gold to iron and nickel.
For instance, one future target of exploration by NASA—an asteroid dubbed 16 Psyche, orbiting the sun in the asteroid belt between Mars and Jupiter—is worth an estimated $10,000 quadrillion. It’s a number so mind-bogglingly big that it would crash the global economy, if someone ever figured out how to tow it back to Earth without literally crashing it into the planet.
Living Off the Land
Space mining isn’t just about getting rich. Many argue that humanity’s ability to extract resources in space, especially water that can be refined into rocket fuel, will be a key technology to extend our reach beyond near-Earth space.
The presence of frozen water around the frigid polar regions of the moon, for example, represents an invaluable source to power future deep-space missions. Splitting H20 into its component elements of hydrogen and oxygen would provide a nearly inexhaustible source of rocket fuel. Today, it costs $10,000 to put a pound of payload in Earth orbit, according to NASA.
Until more advanced rocket technology is developed, the moon looks to be the best bet for serving as the launching pad to Mars and beyond.
Moon Versus Asteroid
However, Lewicki notes that despite the moon’s proximity and our more intimate familiarity with its pockmarked surface, that doesn’t mean a lunar mission to extract resources is any easier than a multi-year journey to a fast-moving asteroid.
For one thing, fighting gravity to and from the moon is no easy feat, as the moon has a significantly stronger gravitational field than an asteroid. Another challenge is that the frozen water is located in permanently shadowed lunar craters, meaning space miners can’t rely on solar-powered equipment, but on some sort of external energy source.
And then there’s the fact that moon craters might just be the coldest places in the solar system. NASA’s Lunar Reconnaissance Orbiter found temperatures plummeted as low as 26 Kelvin, or more than minus 400 degrees Fahrenheit. In comparison, the coldest temperatures on Earth have been recorded near the South Pole in Antarctica—about minus 148 degrees F.
“We don’t operate machines in that kind of thermal environment,” Lewicki said of the extreme temperatures detected in the permanent dark regions of the moon. “Antarctica would be a balmy desert island compared to a lunar polar crater.”
Of course, no one knows quite what awaits us in the asteroid belt. Answers may soon be forthcoming. Last week, the Japan Aerospace Exploration Agency landed two small, hopping rovers on an asteroid called Ryugu. Meanwhile, NASA hopes to retrieve a sample from the near-Earth asteroid Bennu when its OSIRIS-REx mission makes contact at the end of this year.
No Bucks, No Buck Rogers
Visionaries like Elon Musk and Jeff Bezos talk about colonies on Mars, with millions of people living and working in space. The reality is that there’s probably a reason Buck Rogers was set in the 25th century: It’s going to take a lot of money and a lot of time to realize those sci-fi visions.
Or, as Lewicki put it: “No bucks, no Buck Rogers.”
The cost of operating in outer space can be prohibitive. Planetary Resources itself is grappling with raising additional funding, with reports this year about layoffs and even a possible auction of company assets.
Still, Lewicki is confident that despite economic and technical challenges, humanity will someday exceed even the boldest dreamers—skyscrapers on the moon, interplanetary trips to Mars—as judged against today’s engineering marvels.
“What we’re doing is going to be very hard, very painful, and almost certainly worth it,” he said. “Who would have thought that there would be a job for a space miner that you could go to school for, even just five or ten years ago. Things move quickly.”
Image Credit: M-SUR / Shutterstock.com Continue reading →
Posted in Human Robots