Five Notable Advances in Artificial Intelligence of 2018
Photo Source: Wikimedia Commons
With the year now coming to a close, it makes for a perfect time to reflect on the many breakthroughs and advancements in artificial intelligence that occurred in 2018. The affects of artificial intelligence continue to expand their influence in almost every field from commerce to science and law in society. Many of the more worrisome aspects of artificial intelligence have already surfaced in science fiction from the rogue computer Hal in Stanley Kubrick’s 2001: A Space Odyssey to a dystopian future where individuals get accused and tried for crimes before they commit them in The Minority Report. The following short list of artificial intelligence advances from 2018 highlight some exciting and, in some cases, worrisome breakthroughs in artificial intelligence this year.
1; Echoes of 2001: A Space Odyssey occurred this year when the robot, CIMON, appeared to get peevish with an astronaut in a video released on November 30, 2018, by the European Space Agency. CIMON stands for Crew Interactive MObile CompanioN, which represents the new robot deployed on the International Space Station (ISS). The European Space Agency in collaboration with Airbus, the German Aerospace Center, and IBM’s Watson recently launched the first artificially intelligent robot into space and deployed it on the International Space Station. CIMON plays the dual roles of smart, interactive assistant and social companion capable of conversation. The interesting component of the interaction between the astronaut and CIMON came when the robot appeared to get sensitive to the astronaut’s comments. The robot tells the astronaut Alexander Gerst, “Be nice!” and eventually “Don’t be mean!” The robot’s artificial intelligence appeared to be sensitive to the comments and commands of the astronaut. Clearly, the artificial intelligence picked up something in the conversation that caused it to try to correct the behavior of the astronaut. Artificial intelligence in the early stages of development will take on the characteristics of the people that train it, which should inform the researchers developing it to take extra care to impart the best of humanity as the machines get smarter and smarter.
2. The frightening aspect of The Minority Report stems from the assignment of guilt for crimes not yet committed, and it appears to have begun in reality in Britain. Two thousand eighteen unveiled a new system in England that uses unprecedented surveillance combined with artificial intelligence to anticipate crimes and attempt interventions before the crimes can occur. The author, Kit Klarenberg, for Sputniknews, described a total surveillance system as “a vast store of gathered intelligence — as well as all information held on various other police data repositories. This includes sensitive information on victims of crime, individuals who've been cleared of any wrongdoing, and other content unrelated to criminal activities — such as driver and vehicle records, around 55 million in number apiece.” Authorities call the system the National Data Analytics Solution, and they insist that it will not be used to prosecute pre-crimes. Instead, the system will provide tools for intervention for the most likely to commit a crime. National Data Analytics Solution still sure sounds like a step towards The Minority Report.
3. Doctors depend on many tests to diagnose disease, and the interpretation of some tests require the trained eye of a specialist such as a radiologist to identify conditions in the body. In cases such as a stroke, fast early detection can help prevent catastrophic long-term effects with the right intervention. Artificial intelligence stands to become an invaluable physician’s assistant to interpret test results more quickly and accurately than ever before. Senior author Eric Oermann, MD, instructor in the department of neurosurgery at the Icahn School of Medicine at Mount Sinai in New York, described in an article in the journal Nature Medicine (Volume 24, pages1337–1341, 2018) a new artificial intelligence-based system that can detect disease from CT scans. CT scans, or computed tomography scans, use x-rays and computer imaging to look inside the body to find disease. Michael Walter for Radiology News quotes Dr. Oermann as saying, “With a total processing and interpretation time of 1.2 seconds, such a triage system can alert physicians to a critical finding that may otherwise remain in a queue for minutes to hours. Moreover, he notes. “On average, the AI was able to preprocess an image, read it and alert a radiologist—when necessary—approximately 150 times faster than it would take a physician to interpret the exam.” Such an improvement in processing CT scans would help physicians move more quickly to help patients with interventions that may reduce the affects of stroke and other diseases, which highlights a fascinating and beneficial aspect of artificial intelligence.
4. The future of fast food may be in the “hands” of robots powered by artificial intelligence. In 2018, Flippy the robot went live at Caliburger in Pasadena, CA as a short order cook that flips hamburgers and with the help of artificial intelligence. According to Miso Robotics, the company that manufactures Flippy, the robot uses thermal sensors, 3D vision, and normal vision to automatically detect when raw burger patties get placed on the grill. Additionally, with the help of artificial intelligence, Flippy monitors each patty in real-time throughout the cooking process, switches between a spatula for raw meat and one for cooked meat, cleans spatulas while cooking, and wipes the surface of the grill with a scraper. Flippy works with people to make hamburgers and serves to make sure each hamburger gets cooked the appropriate amount of time to be safe and consistent. With minimum wages rising, restaurant owners may look at using smart robots to make their food preparation more consistent and inexpensive. However, we must consider the implications of such changes on the lives of workers and our dining experience before rushing in to have a robot short order cook.
5. For the most part, robots have not fared well walking across uneaven terrain that people and animals have adapted to over thousands of years. Robots tend to stumble and fall when trying to walk across anything but a smooth flat surface. However, research in robot mobility has vastly improved with the inclusion of artificial intelligence to help robots move in uneven terrain. Two thousand eighteen witnessed remarkable improvements in mobility. In “Humanoid Robot Atlas Can Now Do Parkour, and That's Not at All Terrifying,” Mindy Weisberger, Senior Writer for Live Science, describes the latest advances in the mobility and agility of the robot called Atlas developed by Boston Dynamics. She writes, “Atlas demonstrates mobility that is uncannily human: recovering after being shoved, performing backflips, jogging over a grassy field and practicing robot parkour. Parkour for the unfamiliar, including me, describes the activity or moving through an uneven area, rapidly using skills such as running, jumping, and climbing. The days of leaving robots at the roadside appear to be ending as Atlas conquers more challenging territory and will soon follow humans in much more challenging terrain. Watch the following video to see Atlas in action.
Two thousand eighteen brought many artificial intelligence breakthroughs across every aspect of industry and society. The surging developments with artificial intelligence promise to make 2019 even more impactful. As consumers and citizens we should take advantage of developments in artificial intelligence, but we should also be critical and vocal about developments that may not be beneficial for society. Have a Happy New Year!
Dr. Smith’s career in scientific and information research spans the areas of bioinformatics, artificial intelligence, toxicology, and chemistry. He has published a number of peer-reviewed scientific papers. He has worked over the past seventeen years developing advanced analytics, machine learning, and knowledge management tools to enable research and support high level decision making. Tim completed his Ph.D. in Toxicology at Cornell University and a Bachelor of Science in chemistry from the University of Washington.
You can buy his book on Amazon in paperback and in kindle format here.