Machine learning classifiers are known to be vulnerable to inputs maliciously constructed by adversaries to force misclassification. Such adversarial examples have been extensively studied in the context of computer vision applications. In this work, we show adversarial attacks are also effective when targeting neural network policies in reinforcement learning. Specifically, we show existing adversarial example crafting techniques can be used to significantly degrade test-time performance of trained policies. Our threat model considers adversaries capable of introducing small perturbations to the raw input of the policy. We characterize the degree of vulnerability across tasks and training algorithms, for a subclass of adversarial-example attacks in white-box and black-box settings. Regardless of the learned task or training algorithm, we observe a significant drop in performance, even with small adversarial perturbations that do not interfere with human perception. Videos are available at http://rll.berkeley.edu/adversarial.

The past couple of years have been a real cringe-y time to be an AI researcher. Just imagine a whole bunch of famous technologists and top-serious science authorities all suddenly taking aim at your field of research as a clear and present threat to the very survival of the species. All you want to do is predict appropriate emoji use based on textual analyses and here's Elon Musk saying this thing he doesn't really seem to know much about is the actual apocalypse. It's not that computer scientists haven't argued against AI hype, but an academic you've never heard of (all of them?) pitching the headline "AI is hard" is at a disadvantage to the famous person whose job description largely centers around making big public pronouncements. This month that academic is Alan Bundy, a professor of automated reasoning at the University of Edinburgh in Scotland, who argues in the Communications of the ACM that there is a real AI threat, but it's not human-like machine intelligence gone amok.

Former NASA engineer Mark Moore will now be working on Uber's flying car project, Uber Elevate. SAN FRANCISCO -- George Jetson, your ride is on its way. Uber has just hired a NASA expert to build out its vision for flying cars Monday. Mark Moore, a 30-year veteran of the space agency with expertise in using electric motors to get a vehicle airborne, will help the ride-hailing giant execute on an expansive white paper it released last fall on developing VTOL (vertical take-off and landing) vehicles. "Uber continues to see its role as a catalyst to the growing developing VTOL ecosystem," Nikhil Goel, Uber's head of product for advanced programs, said in a statement.

In the race to make a commercially available flying car, Uber has hired Mark Moore, a talented Nasa aircraft engineer. Moore inspired Google's co-founder, Larry Page, to get involved in flying car projects. Uber now hopes that with Moore's help, it will be able to have flying cars ready in'one to three years.' In the race to make a commercially available flying car, Uber has hired Mark Moore, a talented Nasa aircraft engineer. Moore predicts it wont be long before we see several well-engineered flying cars, estimating it will only be'one to three years' Uber has a plan to launch a fleet of electric aircraft for on-demand aviation. The plan, called Uber Elevate, is set to design these'cars' to take off and land vertically, travel 100 miles on a single charge and has set 2021 as the ready date - although Mr Moore predicts it could be even sooner.

The Cambridge Centre for Risk Studies in collaboration with the United Nations programme on Journalism and Public Information (UNICRI) present a workshop on the Risks and Benefits of Artificial Intelligence and Robotics. From sensing, finance, medicine, transportation and security, a technological revolution is taking place. Artificial Intelligence (AI) has been a feature of science fiction for almost a century, but it is only in more recent years that the prospect of autonomous robotics and artificially intelligent systems has really become viable. While this will potentially provide great opportunities, these developments are likely to have significant impacts upon the very functioning of society, posing practical, ethical, legal and security challenges – much of which is as of yet not fully appreciated or understood. The media and other sources of public information are central in ensuring that citizens and institutions have a realistic and balanced understanding of such technologies.

Today, Mark takes on that ongoing tension of artificial intelligence and jobs, looking at how things are looking for jobs in Government administration, healthcare, and education. Public sector jobs could be under threat by robots in the next 15 years, according to a new report by The Reform think-tank. As technology is advancing at an unprecedented rate, the report explains, there is a growing opportunity to improve efficiency to the tune of billions of pounds by engaging robots to undertake certain tasks. The tasks that these bots can undertake, however, may have ramifications for civil servants' jobs. The co-author of the report, Alexander Hitchcock, has put this caveat on The Reform's think-tank: "Such a rapid advance in the use of technology may seem controversial, and any job losses must be handled sensitively. But the result will be public services that are better, safer, smarter, and more affordable."

Writing in the Economist newspaper on 8 October 2016, US President Barack Obama called out artificial intelligence (AI) as one of several areas where'in recent years we have seen incredible technological advances'.2 Long a backroom area of computer science, AI has captured the popular imagination over the last two years as the range and impact of practical AI applications have expanded at a dizzying pace: a quick search on ft.com for'artificial intelligence and robotics' returned 4 stories from September and October 2014, 16 for the same period in 2015 and 54 in 2016. AI is one of several areas of digital innovation that are all both developing increasingly rapidly and interacting with each other in ways whose consequences are challenging to foresee. A useful portmanteau for these changes is the'fourth industrial revolution'. After steam, electricity and computing, this is the term coined3 by Davos founder Klaus Schwab for the deep digital transformation now upon us.

As our Supreme Court weighed the First Amendment implications of brutal video games back in 2011, Justice Samuel Alito cut in with a sarcastic jab: "Well, I think what Justice Scalia wants to know is what James Madison thought about video games. Laurence H. Tribe is the Carl M. Loeb University Professor and professor of constitutional law at Harvard University. Tribe and Matz are the co-authors of Uncertain Justice: The Roberts Court and the Constitution." This wasn't the first time that scientific advances had divided these super-conservative justices--and that speaks to a crucial point. While the confirmation hearings for Judge Neil Gorsuch will involve familiar debates over how to read the Constitution, judicial orientations toward new technology can scramble the fields in surprising ways.

With so much discussion over foreign policy and immigration concerns from Muslim-majority nations after the inauguration of Donald Trump, there might be some confusion about how the U.S. is fighting terrorism in the Middle East. The U.S. may not be in a direct war with anyone other than the Islamic State, also known as ISIS, but there is still a military presence in multiple countries carried over from previous administrations. The number of combat troops has dipped due to drone warfare, but President Trump, who campaigned on being tougher on ISIS, has said he would be willing to send up to 30,000 troops to Iraq and Syria. However, he inherited a military presence in not just those two countries, but other hotspots, as well. Just days after the Sept. 11 attacks, Congress and President George W. Bush authorized the use of military force to overthrow the Taliban.

The patter of automated machinery fills the air inside wire-basket manufacturer Marlin Steel's bustling factory in a rugged industrial section of this city. Maxi Cifarelli, 25, of Baltimore, peers through safety goggles at a flat screen, her left knee bent and heel resting on her chair. Two years after earning a fine arts degree from Towson University with a specialty in interdisciplinary object design, she now spends her work days working with a personality-free machine with a name to match: a computer numerical control, or CNC, router. With automation poised to sweep through the economy, some fear that it will kill more jobs than it creates. But Cifarelli's experience is the opposite. She befriended automation, instead of fighting it, and she has a job because of it.