Tag Archives: answer
#434759 To Be Ethical, AI Must Become ...
As over-hyped as artificial intelligence is—everyone’s talking about it, few fully understand it, it might leave us all unemployed but also solve all the world’s problems—its list of accomplishments is growing. AI can now write realistic-sounding text, give a debating champ a run for his money, diagnose illnesses, and generate fake human faces—among much more.
After training these systems on massive datasets, their creators essentially just let them do their thing to arrive at certain conclusions or outcomes. The problem is that more often than not, even the creators don’t know exactly why they’ve arrived at those conclusions or outcomes. There’s no easy way to trace a machine learning system’s rationale, so to speak. The further we let AI go down this opaque path, the more likely we are to end up somewhere we don’t want to be—and may not be able to come back from.
In a panel at the South by Southwest interactive festival last week titled “Ethics and AI: How to plan for the unpredictable,” experts in the field shared their thoughts on building more transparent, explainable, and accountable AI systems.
Not New, but Different
Ryan Welsh, founder and director of explainable AI startup Kyndi, pointed out that having knowledge-based systems perform advanced tasks isn’t new; he cited logistical, scheduling, and tax software as examples. What’s new is the learning component, our inability to trace how that learning occurs, and the ethical implications that could result.
“Now we have these systems that are learning from data, and we’re trying to understand why they’re arriving at certain outcomes,” Welsh said. “We’ve never actually had this broad society discussion about ethics in those scenarios.”
Rather than continuing to build AIs with opaque inner workings, engineers must start focusing on explainability, which Welsh broke down into three subcategories. Transparency and interpretability come first, and refer to being able to find the units of high influence in a machine learning network, as well as the weights of those units and how they map to specific data and outputs.
Then there’s provenance: knowing where something comes from. In an ideal scenario, for example, Open AI’s new text generator would be able to generate citations in its text that reference academic (and human-created) papers or studies.
Explainability itself is the highest and final bar and refers to a system’s ability to explain itself in natural language to the average user by being able to say, “I generated this output because x, y, z.”
“Humans are unique in our ability and our desire to ask why,” said Josh Marcuse, executive director of the Defense Innovation Board, which advises Department of Defense senior leaders on innovation. “The reason we want explanations from people is so we can understand their belief system and see if we agree with it and want to continue to work with them.”
Similarly, we need to have the ability to interrogate AIs.
Two Types of Thinking
Welsh explained that one big barrier standing in the way of explainability is the tension between the deep learning community and the symbolic AI community, which see themselves as two different paradigms and historically haven’t collaborated much.
Symbolic or classical AI focuses on concepts and rules, while deep learning is centered around perceptions. In human thought this is the difference between, for example, deciding to pass a soccer ball to a teammate who is open (you make the decision because conceptually you know that only open players can receive passes), and registering that the ball is at your feet when someone else passes it to you (you’re taking in information without making a decision about it).
“Symbolic AI has abstractions and representation based on logic that’s more humanly comprehensible,” Welsh said. To truly mimic human thinking, AI needs to be able to both perceive information and conceptualize it. An example of perception (deep learning) in an AI is recognizing numbers within an image, while conceptualization (symbolic learning) would give those numbers a hierarchical order and extract rules from the hierachy (4 is greater than 3, and 5 is greater than 4, therefore 5 is also greater than 3).
Explainability comes in when the system can say, “I saw a, b, and c, and based on that decided x, y, or z.” DeepMind and others have recently published papers emphasizing the need to fuse the two paradigms together.
Implications Across Industries
One of the most prominent fields where AI ethics will come into play, and where the transparency and accountability of AI systems will be crucial, is defense. Marcuse said, “We’re accountable beings, and we’re responsible for the choices we make. Bringing in tech or AI to a battlefield doesn’t strip away that meaning and accountability.”
In fact, he added, rather than worrying about how AI might degrade human values, people should be asking how the tech could be used to help us make better moral choices.
It’s also important not to conflate AI with autonomy—a worst-case scenario that springs to mind is an intelligent destructive machine on a rampage. But in fact, Marcuse said, in the defense space, “We have autonomous systems today that don’t rely on AI, and most of the AI systems we’re contemplating won’t be autonomous.”
The US Department of Defense released its 2018 artificial intelligence strategy last month. It includes developing a robust and transparent set of principles for defense AI, investing in research and development for AI that’s reliable and secure, continuing to fund research in explainability, advocating for a global set of military AI guidelines, and finding ways to use AI to reduce the risk of civilian casualties and other collateral damage.
Though these were designed with defense-specific aims in mind, Marcuse said, their implications extend across industries. “The defense community thinks of their problems as being unique, that no one deals with the stakes and complexity we deal with. That’s just wrong,” he said. Making high-stakes decisions with technology is widespread; safety-critical systems are key to aviation, medicine, and self-driving cars, to name a few.
Marcuse believes the Department of Defense can invest in AI safety in a way that has far-reaching benefits. “We all depend on technology to keep us alive and safe, and no one wants machines to harm us,” he said.
A Creation Superior to Its Creator
That said, we’ve come to expect technology to meet our needs in just the way we want, all the time—servers must never be down, GPS had better not take us on a longer route, Google must always produce the answer we’re looking for.
With AI, though, our expectations of perfection may be less reasonable.
“Right now we’re holding machines to superhuman standards,” Marcuse said. “We expect them to be perfect and infallible.” Take self-driving cars. They’re conceived of, built by, and programmed by people, and people as a whole generally aren’t great drivers—just look at traffic accident death rates to confirm that. But the few times self-driving cars have had fatal accidents, there’s been an ensuing uproar and backlash against the industry, as well as talk of implementing more restrictive regulations.
This can be extrapolated to ethics more generally. We as humans have the ability to explain our decisions, but many of us aren’t very good at doing so. As Marcuse put it, “People are emotional, they confabulate, they lie, they’re full of unconscious motivations. They don’t pass the explainability test.”
Why, then, should explainability be the standard for AI?
Even if humans aren’t good at explaining our choices, at least we can try, and we can answer questions that probe at our decision-making process. A deep learning system can’t do this yet, so working towards being able to identify which input data the systems are triggering on to make decisions—even if the decisions and the process aren’t perfect—is the direction we need to head.
Image Credit: a-image / Shutterstock.com Continue reading
Posted in Human Robots
#434701 3 Practical Solutions to Offset ...
In recent years, the media has sounded the alarm about mass job loss to automation and robotics—some studies predict that up to 50 percent of current jobs or tasks could be automated in coming decades. While this topic has received significant attention, much of the press focuses on potential problems without proposing realistic solutions or considering new opportunities.
The economic impacts of AI, robotics, and automation are complex topics that require a more comprehensive perspective to understand. Is universal basic income, for example, the answer? Many believe so, and there are a number of experiments in progress. But it’s only one strategy, and without a sustainable funding source, universal basic income may not be practical.
As automation continues to accelerate, we’ll need a multi-pronged approach to ease the transition. In short, we need to update broad socioeconomic strategies for a new century of rapid progress. How, then, do we plan practical solutions to support these new strategies?
Take history as a rough guide to the future. Looking back, technology revolutions have three themes in common.
First, past revolutions each produced profound benefits to productivity, increasing human welfare. Second, technological innovation and technology diffusion have accelerated over time, each iteration placing more strain on the human ability to adapt. And third, machines have gradually replaced more elements of human work, with human societies adapting by moving into new forms of work—from agriculture to manufacturing to service, for example.
Public and private solutions, therefore, need to be developed to address each of these three components of change. Let’s explore some practical solutions for each in turn.
Figure 1. Technology’s structural impacts in the 21st century. Refer to Appendix I for quantitative charts and technological examples corresponding to the numbers (1-22) in each slice.
Solution 1: Capture New Opportunities Through Aggressive Investment
The rapid emergence of new technology promises a bounty of opportunity for the twenty-first century’s economic winners. This technological arms race is shaping up to be a global affair, and the winners will be determined in part by who is able to build the future economy fastest and most effectively. Both the private and public sectors have a role to play in stimulating growth.
At the country level, several nations have created competitive strategies to promote research and development investments as automation technologies become more mature.
Germany and China have two of the most notable growth strategies. Germany’s Industrie 4.0 plan targets a 50 percent increase in manufacturing productivity via digital initiatives, while halving the resources required. China’s Made in China 2025 national strategy sets ambitious targets and provides subsidies for domestic innovation and production. It also includes building new concept cities, investing in robotics capabilities, and subsidizing high-tech acquisitions abroad to become the leader in certain high-tech industries. For China, specifically, tech innovation is driven partially by a fear that technology will disrupt social structures and government control.
Such opportunities are not limited to existing economic powers. Estonia’s progress after the breakup of the Soviet Union is a good case study in transitioning to a digital economy. The nation rapidly implemented capitalistic reforms and transformed itself into a technology-centric economy in preparation for a massive tech disruption. Internet access was declared a right in 2000, and the country’s classrooms were outfitted for a digital economy, with coding as a core educational requirement starting at kindergarten. Internet broadband speeds in Estonia are among the fastest in the world. Accordingly, the World Bank now ranks Estonia as a high-income country.
Solution 2: Address Increased Rate of Change With More Nimble Education Systems
Education and training are currently not set for the speed of change in the modern economy. Schools are still based on a one-time education model, with school providing the foundation for a single lifelong career. With content becoming obsolete faster and rapidly escalating costs, this system may be unsustainable in the future. To help workers more smoothly transition from one job into another, for example, we need to make education a more nimble, lifelong endeavor.
Primary and university education may still have a role in training foundational thinking and general education, but it will be necessary to curtail rising price of tuition and increase accessibility. Massive open online courses (MooCs) and open-enrollment platforms are early demonstrations of what the future of general education may look like: cheap, effective, and flexible.
Georgia Tech’s online Engineering Master’s program (a fraction of the cost of residential tuition) is an early example in making university education more broadly available. Similarly, nanodegrees or microcredentials provided by online education platforms such as Udacity and Coursera can be used for mid-career adjustments at low cost. AI itself may be deployed to supplement the learning process, with applications such as AI-enhanced tutorials or personalized content recommendations backed by machine learning. Recent developments in neuroscience research could optimize this experience by perfectly tailoring content and delivery to the learner’s brain to maximize retention.
Finally, companies looking for more customized skills may take a larger role in education, providing on-the-job training for specific capabilities. One potential model involves partnering with community colleges to create apprenticeship-style learning, where students work part-time in parallel with their education. Siemens has pioneered such a model in four states and is developing a playbook for other companies to do the same.
Solution 3: Enhance Social Safety Nets to Smooth Automation Impacts
If predicted job losses to automation come to fruition, modernizing existing social safety nets will increasingly become a priority. While the issue of safety nets can become quickly politicized, it is worth noting that each prior technological revolution has come with corresponding changes to the social contract (see below).
The evolving social contract (U.S. examples)
– 1842 | Right to strike
– 1924 | Abolish child labor
– 1935 | Right to unionize
– 1938 | 40-hour work week
– 1962, 1974 | Trade adjustment assistance
– 1964 | Pay discrimination prohibited
– 1970 | Health and safety laws
– 21st century | AI and automation adjustment assistance?
Figure 2. Labor laws have historically adjusted as technology and society progressed
Solutions like universal basic income (no-strings-attached monthly payout to all citizens) are appealing in concept, but somewhat difficult to implement as a first measure in countries such as the US or Japan that already have high debt. Additionally, universal basic income may create dis-incentives to stay in the labor force. A similar cautionary tale in program design was the Trade Adjustment Assistance (TAA), which was designed to protect industries and workers from import competition shocks from globalization, but is viewed as a missed opportunity due to insufficient coverage.
A near-term solution could come in the form of graduated wage insurance (compensation for those forced to take a lower-paying job), including health insurance subsidies to individuals directly impacted by automation, with incentives to return to the workforce quickly. Another topic to tackle is geographic mismatch between workers and jobs, which can be addressed by mobility assistance. Lastly, a training stipend can be issued to individuals as means to upskill.
Policymakers can intervene to reverse recent historical trends that have shifted incomes from labor to capital owners. The balance could be shifted back to labor by placing higher taxes on capital—an example is the recently proposed “robot tax” where the taxation would be on the work rather than the individual executing it. That is, if a self-driving car performs the task that formerly was done by a human, the rideshare company will still pay the tax as if a human was driving.
Other solutions may involve distribution of work. Some countries, such as France and Sweden, have experimented with redistributing working hours. The idea is to cap weekly hours, with the goal of having more people employed and work more evenly spread. So far these programs have had mixed results, with lower unemployment but high costs to taxpayers, but are potential models that can continue to be tested.
We cannot stop growth, nor should we. With the roles in response to this evolution shifting, so should the social contract between the stakeholders. Government will continue to play a critical role as a stabilizing “thumb” in the invisible hand of capitalism, regulating and cushioning against extreme volatility, particularly in labor markets.
However, we already see business leaders taking on some of the role traditionally played by government—thinking about measures to remedy risks of climate change or economic proposals to combat unemployment—in part because of greater agility in adapting to change. Cross-disciplinary collaboration and creative solutions from all parties will be critical in crafting the future economy.
Note: The full paper this article is based on is available here.
Image Credit: Dmitry Kalinovsky / Shutterstock.com Continue reading
Posted in Human Robots
#434559 Can AI Tell the Difference Between a ...
Scarcely a day goes by without another headline about neural networks: some new task that deep learning algorithms can excel at, approaching or even surpassing human competence. As the application of this approach to computer vision has continued to improve, with algorithms capable of specialized recognition tasks like those found in medicine, the software is getting closer to widespread commercial use—for example, in self-driving cars. Our ability to recognize patterns is a huge part of human intelligence: if this can be done faster by machines, the consequences will be profound.
Yet, as ever with algorithms, there are deep concerns about their reliability, especially when we don’t know precisely how they work. State-of-the-art neural networks will confidently—and incorrectly—classify images that look like television static or abstract art as real-world objects like school-buses or armadillos. Specific algorithms could be targeted by “adversarial examples,” where adding an imperceptible amount of noise to an image can cause an algorithm to completely mistake one object for another. Machine learning experts enjoy constructing these images to trick advanced software, but if a self-driving car could be fooled by a few stickers, it might not be so fun for the passengers.
These difficulties are hard to smooth out in large part because we don’t have a great intuition for how these neural networks “see” and “recognize” objects. The main insight analyzing a trained network itself can give us is a series of statistical weights, associating certain groups of points with certain objects: this can be very difficult to interpret.
Now, new research from UCLA, published in the journal PLOS Computational Biology, is testing neural networks to understand the limits of their vision and the differences between computer vision and human vision. Nicholas Baker, Hongjing Lu, and Philip J. Kellman of UCLA, alongside Gennady Erlikhman of the University of Nevada, tested a deep convolutional neural network called VGG-19. This is state-of-the-art technology that is already outperforming humans on standardized tests like the ImageNet Large Scale Visual Recognition Challenge.
They found that, while humans tend to classify objects based on their overall (global) shape, deep neural networks are far more sensitive to the textures of objects, including local color gradients and the distribution of points on the object. This result helps explain why neural networks in image recognition make mistakes that no human ever would—and could allow for better designs in the future.
In the first experiment, a neural network was trained to sort images into 1 of 1,000 different categories. It was then presented with silhouettes of these images: all of the local information was lost, while only the outline of the object remained. Ordinarily, the trained neural net was capable of recognizing these objects, assigning more than 90% probability to the correct classification. Studying silhouettes, this dropped to 10%. While human observers could nearly always produce correct shape labels, the neural networks appeared almost insensitive to the overall shape of the images. On average, the correct object was ranked as the 209th most likely solution by the neural network, even though the overall shapes were an exact match.
A particularly striking example arose when they tried to get the neural networks to classify glass figurines of objects they could already recognize. While you or I might find it easy to identify a glass model of an otter or a polar bear, the neural network classified them as “oxygen mask” and “can opener” respectively. By presenting glass figurines, where the texture information that neural networks relied on for classifying objects is lost, the neural network was unable to recognize the objects by shape alone. The neural network was similarly hopeless at classifying objects based on drawings of their outline.
If you got one of these right, you’re better than state-of-the-art image recognition software. Image Credit: Nicholas Baker, Hongjing Lu, Gennady Erlikhman, Philip J. Kelman. “Deep convolutional networks do not classify based on global object shape.” Plos Computational Biology. 12/7/18. / CC BY 4.0
When the neural network was explicitly trained to recognize object silhouettes—given no information in the training data aside from the object outlines—the researchers found that slight distortions or “ripples” to the contour of the image were again enough to fool the AI, while humans paid them no mind.
The fact that neural networks seem to be insensitive to the overall shape of an object—relying instead on statistical similarities between local distributions of points—suggests a further experiment. What if you scrambled the images so that the overall shape was lost but local features were preserved? It turns out that the neural networks are far better and faster at recognizing scrambled versions of objects than outlines, even when humans struggle. Students could classify only 37% of the scrambled objects, while the neural network succeeded 83% of the time.
Humans vastly outperform machines at classifying object (a) as a bear, while the machine learning algorithm has few problems classifying the bear in figure (b). Image Credit: Nicholas Baker, Hongjing Lu, Gennady Erlikhman, Philip J. Kelman. “Deep convolutional networks do not classify based on global object shape.” Plos Computational Biology. 12/7/18. / CC BY 4.0
“This study shows these systems get the right answer in the images they were trained on without considering shape,” Kellman said. “For humans, overall shape is primary for object recognition, and identifying images by overall shape doesn’t seem to be in these deep learning systems at all.”
Naively, one might expect that—as the many layers of a neural network are modeled on connections between neurons in the brain and resemble the visual cortex specifically—the way computer vision operates must necessarily be similar to human vision. But this kind of research shows that, while the fundamental architecture might resemble that of the human brain, the resulting “mind” operates very differently.
Researchers can, increasingly, observe how the “neurons” in neural networks light up when exposed to stimuli and compare it to how biological systems respond to the same stimuli. Perhaps someday it might be possible to use these comparisons to understand how neural networks are “thinking” and how those responses differ from humans.
But, as yet, it takes a more experimental psychology to probe how neural networks and artificial intelligence algorithms perceive the world. The tests employed against the neural network are closer to how scientists might try to understand the senses of an animal or the developing brain of a young child rather than a piece of software.
By combining this experimental psychology with new neural network designs or error-correction techniques, it may be possible to make them even more reliable. Yet this research illustrates just how much we still don’t understand about the algorithms we’re creating and using: how they tick, how they make decisions, and how they’re different from us. As they play an ever-greater role in society, understanding the psychology of neural networks will be crucial if we want to use them wisely and effectively—and not end up missing the woods for the trees.
Image Credit: Irvan Pratama / Shutterstock.com Continue reading
Posted in Human Robots