Before autonomous trucks and taxis hit the road, manufacturers will need to solve problems far more complex than collision avoidance and navigation (see "10 Breakthrough Technologies 2017: Self-Driving Trucks"). These vehicles will have to anticipate and defend against a full spectrum of malicious attackers wielding both traditional cyberattacks and a new generation of attacks based on so-called adversarial machine learning (see "AI Fight Club Could Help Save Us from a Future of Super-Smart Cyberattacks"). When hackers demonstrated that vehicles on the roads were vulnerable to several specific security threats, automakers responded by recalling and upgrading the firmware of millions of cars. The computer vision and collision avoidance systems under development for autonomous vehicles rely on complex machine-learning algorithms that are not well understood, even by the companies that rely on them (see "The Dark Secret at the Heart of AI").
When it comes to digital assistants like Amazon's Alexa, my four-year-old niece Hannah Metz is an early adopter. "Alexa, play'It's Raining Tacos,'" she commanded on a recent sunny afternoon, and the voice-controlled helper immediately complied, blasting through its speaker a confection of a song with lines like "It's raining tacos from out of the sky" and "Yum, yum, yum, yum, yumidy yum." These things are most popular among people age 25 to 34, which includes a ton of parents of young children and parents-to-be. Her interest in her digital assistant jibes with some findings in a recent MIT study, where researchers looked at how children ages three to 10 interacted with Alexa, Google Home, a tiny game-playing robot called Cozmo, and a smartphone app called Julie Chatbot.
Online ordering and home delivery: Online ordering is hugely convenient. Robot workforce: Factories increasingly have fewer and fewer human workers, which means no personalities to deal with, no agitating for overtime, and no illnesses. Big data: Improvements and innovations in crunching massive amounts of data mean that patterns can be recognized in our behavior where they weren't seen previously. Automated high-speed stock buying and selling: A machine crunching huge amounts of data can spot trends and patterns quickly and act on them faster than a person can.
Google, Apple, Samsung, and Microsoft are each putting thousands of researchers and business specialists to work trying to create irresistible versions of easy-to-use devices that we can talk with. Now the new user interfaces are bending to us," observes Ahmed Bouzid, the chief executive officer of Witlingo, which builds voice-driven apps of all sorts for banks, universities, law firms, and others. To that end, Amazon is encouraging independent developers to build new services on the platform, just as Apple has long done with app developers. During a meeting at the company's Cambridge offices, I asked Alexa's head scientist, Rohit Prasad, why he needs so many people--and when his research team might be fully built out.
Both the DeepMind and CMU approaches use deep reinforcement learning, popularized by DeepMind's Atari-playing AI. A neural network is fed raw pixel data from a virtual environment and uses rewards, like points in a computer game, to learn by trial and error (see "10 Breakthrough Technologies 2017: Reinforcement Learning"). By running through millions of training scenarios at accelerated speeds, both AI programs learned to associate words with particular objects and characteristics, which let them follow the commands. The millions of training runs required means Domingos is not convinced pure deep reinforcement learning will ever crack the real world.
The race to build mass-market autonomous cars is creating big demand for laser sensors that help vehicles map their surroundings. Most driverless cars make use of lidar sensors, which bounce laser beams off nearby objects to create 3-D maps of their surroundings. Each beam is separated by an angle of 0.4 (smaller angles between beams equal higher resolution), with a range of 120 meters. Austin Russell, the CEO of lidar startup Luminar, says his company actively chose not to use solid-state hardware in its sensors, because it believes that while mechanically steering a beam is more expensive, it currently provides more finely detailed images that are critical for safe driving.
According to an array of predictions from tech companies and market researchers, plenty of changes are coming, including many more developments in transportation, lots of people spending time in virtual reality, lab-grown chicken, and, just maybe, male birth control. A report from Bloomberg New Energy Finance forecasts that electric vehicles will make up a little less than 4 percent of all car sales in the U.S. and 5 percent of all car sales in Europe that year, up from 1 percent and 2 percent, respectively, this year. Clean energy has been getting cheaper for a while now, and in the not-so-distant future of 2021, it is expected to finally become cheaper to use renewable power sources like solar and wind rather than coal in a number of key countries. While solar power--which is pricier than the other big renewable source, wind--is already comparable price-wise to coal in countries such as the U.S., Italy, and Germany, a recent report from Bloomberg New Energy Finance predicts that by 2021, solar power will be cheaper than coal in China, India, Mexico, Brazil, and the U.K. That would be bad news for the already declining coal industry, but potentially good news for the environment, especially in China, India, and other countries where air pollution is a constant concern.
In China today, voice assistant technology works by turning a user's voice commands into text and generating a response based on the meaning of the text. They will also have to understand emotions, since humans' decision making is not based solely on logic, notes Jia Jia, an associate professor at Tsinghua University who studies social affective computing. As of the end of 2016, Baidu claimed 665 million monthly active mobile users, and as of March this year, Alibaba had 507 million mobile monthly active users. For example, to train a neural network to understand texts in sports medicine, you could draw upon data from sports and data from medicine.
The system stores a database of potential ditch sites for safe emergency landings, and is able to choose the ideal site based on range, size, type of terrain, reliability, and time or day constraints. It's a much more advanced system than what is currently used in most commercial UAVs, which require a designated "home" point, to which the vehicle will attempt to return in the case of a hardware malfunction or drained battery. Current models are unable to safely ditch if, for example, the remaining battery charge is unable to return the drone to its home point, or if that home point is out of date. Once these remaining technological challenges are solved, Roy believes that Safe2Ditch, or similar systems, could become an FAA-mandated safety standard in UAV manufacturing.
A new competition heralds what is likely to become the future of cybersecurity and cyberwarfare, with offensive and defensive AI algorithms doing battle. "It's a brilliant idea to catalyze research into both fooling deep neural networks and designing deep neural networks that cannot be fooled," says Jeff Clune, an assistant professor at the University of Wyoming who studies the limits of machine learning. Machine learning, and deep learning in particular, is rapidly becoming an indispensable tool in many industries. "Adversarial machine learning is more difficult to study than conventional machine learning--it's hard to tell if your attack is strong or if your defense is actually weak," says Ian Goodfellow, a researcher at Google Brain, a division of Google dedicated to researching and applying machine learning, who organized the contest.