From unprecedented technological leaps to the dangers they present, In simple terms, AI is the intelligence displayed by machines in contrast with the natural intelligence of human beings. Our brains, however, and the realm of the mind are incredibly complex and both are far from being fully understood scientifically. Some experts define human consciousness as having a unique sense of self, combined with the awareness of everything in the surrounding environment. For machines, consciousness would mean setting goals and making choices that are beyond their programming. Even though current applications have their limits, there are a number of experts who believe that we’re not far from that, technologically speaking. In its most popular definition, this is the theoretical point in the future in which AI will surpass human intelligence. As such, there are a number of growing concerns regarding both present and potential AI technology. Number 7 Autonomous VehiclesThe main purpose of a self-driving vehicle is for it to be demonstrably safer than a human-controlled one. Whether this is possible in the real world, without sacrificing human lives, remains a controversial topic. Between 2016 and 2019 there have been four-driver fatalities, each involving the autopilot feature on a Tesla model. A fifth accident involved a pedestrian fatality, caused by a refitted Volvo with an automated system designed by an Uber subsidiary. There are currently five levels of automated driving, of which only four and five are autonomous. Tesla uses a level 2 autopilot, which still requires the driver’s full awareness. There are still ethical questions to be asked, such as whether a car will prioritize the safety of its driver or of other pedestrians' in high-risk scenarios. Self-driving vehicles need to be driven and monitored for millions of more miles, in fully autonomous mode, before an accurate safety comparison to traditional vehicles is even possible. Regardless, many believe autonomous driving will become an alternative and perhaps even a replacement for traditional driving, in the near future. Before we move on, answer this question. What is Skynet? a. A Wireless Internet Provider. A Fictional AI Villains. A Digital Media Outlet. An Air Travel AgencyLet us know what you think in the comments section below and stay tuned to find out the right answer! So far, only certain aspects found in human intelligence have been replicated by artificial neural networks, which are the AI equivalent of the human brain. Applications are thus subjected to something called the AI effect, which occurs when onlookers discount its behavior as being “real intelligence”. As roboticist Rodney Brooks put it “every time we figure out a piece of it, it stops being magical”. That being said, the modern use of AI is pervasive. It includes several high-profile examples such as image recognition, speech recognition, search engines, online assistants, predicting light delays, medical diagnosis, or autonomous vehicles. Number 6 Learning How to LieDeception isn’t a uniquely human trait. In fact, it’s the behavior of squirrels that have inspired some truly fascinating AI developments. A squirrel will purposefully check empty sites for nuts when another squirrel is nearby, thus confusing it about where the food is hidden. Based on this behavior, researchers at the Georgia Institute of Technology have taught a machine how to lie. This was done through a series of hide-and-seek experiments. The robot that was hiding doubled back toa false marker, thus fooling the robot that was seeking. Ronald Arkin, one of the professors involved in the experiment, envisioned military applications for this technology. A robot that’s guarding ammunition or supplies the battlefield can change its patrolling patterns as a means of deception. There are, however, a number of ethical dilemmas and safety concerns. Even Professor Arkin argued that teaching deception to robots and intelligent machines can be dangerous. Stephen Hawking, Bill Gates, and Elon Musk have all argued that AI will eventually grow beyond human control. Should that ever happen, deception is a dangerous tool for AI to possess. Number 5 Smart HomesSmart homes are becoming increasingly more popular and the Internet of Things, or IoT, has been developing rapidly. If you’re not familiar with IoT, just know that it refers to the embedding of data transmission and computing devices in everyday objects. There’ve been several projects to assist elderly people, who were living alone, with systems that monitor electrical and plumbing networks as well as temperature or movement. Introducing advanced AI to the Internet of things may seem like a good idea, but the technology still needs to be perfected. One cautionary tale comes from a prototype house that was designed by Virginia Tech’s Environmental Systems Laboratory. A product of the FutureHAUS initiative, the building was created with a great degree of automation and a number of smart appliances. It burned to the ground after the electrical malfunction of an automated door. Fortunately, no one was hurt. The incident does, nonetheless, serve as a stark reminder of what can go wrong with home AI. Number 4 AI Outperforming HackersComputer hackers are often depicted as the dreaded enemies of security systems, as those people capable of breaking into secret places and finding information that others can’t. AI, however, is starting to outsmart human hackers. And, unlike them, it doesn’t get tired. There are already competitions in which humans develop super-intelligent hacking systems. The AI programs are able to simultaneously protect their performance and functionality while attacking enemies’ vulnerabilities. These systems are typically tasked with fighting“bad hackers”. Concerns mainly address what would happen should this type of technology fall into the wrong hands. Number 3 Military RobotsNumerous artificial intelligence experts and renowned scientists have expressed their concerns regarding the military use of AI. The so-called killer robots might save many lives on the battlefield but are just as likely to attack those on their own side or innocent civilians. This might sound far-fetched but killer robots are one of the concerns expressed by the Bulletin of the Atomic Scientists, the group that operates the Doomsday Clock. Despite repeated warnings from the scientific community, AI is rapidly becoming the crowning jewel as nations struggle for military supremacy. Autonomous drones or advanced detection technologies are seen as particularly valuable assets. With billions being spent on robotics worldwide, every year, the face of war is changing and may ultimately turn against its own creators. Number 2 XiaoiceAs robots are becoming smarter and more human-like in appearance, it would seem our emotions are the final frontier. Soon, that might no longer be the case. In 2014, the East Asian branch of Microsoft-developed an AI called Xiaoice, based on an emotional computing framework. It’s gradually becoming a complete artificial-intelligence system, through its comprehensive application of algorithms, big data, and cloud computing. It has over 660 million registered users and it's mainly popular in China but has expanded to several other countries. Much more than a regular chatbot, Xiaoicecontinuously learns through conversations and interaction. Studies have shown that, on average, humans would rather talk to it than to a regular person. It understands context, responds with human-like emotions, and can adapt the conversation to keep users engaged for hours. In 2017, Xiaoice published the first AI-authored collection of poems called “The Sunshine Lost Windows”. It has released dozens of songs, with similar quality to human singers. Xiaoice is reportedly in a self-growing and self-learning loop that means it will keep getting better. So, what is Skynet? The right answer is b, a fictional AI villain. In the movie “Terminator”, Skynet is anAI system that becomes self-aware after spreading to millions of computers. Its human creators begin to fear its capabilities and attempt to destroy it. A war ensues in which Skynet’s main tool of destruction are cyborgs, known as “terminators”, which it controls as a network-based groupmind. Number 1 Automation of JobsMany experts has argued against the Terminator scenario, claiming that it’s much more likely will kill our jobs rather than kill us. Throughout most of human history, industrial revolutions have opened the gateway to more workplaces. AI might put an end to that and create major technological unemployment. According to Hod Lipson, director of CornellUniversity’s Creative Machines Labs, the threat will stem from the fact that AI “does almost everything better than almost anyone”. In the US alone, there are numerous machines doing the same jobs that humans used to do. Close to 40 million Americans have a workplace that's at risk of becoming automated in the near future. The most affected sectors will be transportation storage, followed by manufacturing and retail. By the turn of the century, it’s expected that more than half of the jobs in the world will be done by AI. There are some, however, likely to survive the AI revolution. These include the work of psychiatrists and other medical professionals, teachers, artists, or church ministers, as most people will likely refuse being preached to by a robot.
0 Comments