Both the artificial intelligence (AI) and robotics fields are expanding rapidly with exciting innovations being built to automate our world. There are many questions about the future of robotics and how robots will integrate and interact with humans. Zhi-Hua Zhou, IEEE Fellow and chair of the IEEE Computational Intelligence Society Data Mining Technical Committee is an IEEE expert from China on robotics, and shares his insight on the future impact of AI and robots.
IEEE Transmitter: How intelligent are robots today and how much smarter will they become?
Zhou: As robotics is a comprehensive subject, it’s difficult to measure “how intelligent robots are.” In some fields, for example, voice recognition and visual recognition, robots today have already obtained similar or even better capabilities compared with human beings. However, in terms of understanding context and responding to complex situations, robots still have a long way to go.
IEEE Transmitter: Will robots be able to learn anything, show an ability to reason or even formulate original ideas?
Zhou: Current intelligent robots are capable of learning and reasoning. They are already equipped with autonomic learning ability through data collection and analysis from surrounding environment. So I think robots have the ability to reason. It’s difficult to define whether they can think independently, in cases of reasoning and formulating ideas.
IEEE Transmitter: How do you think AI will most benefit humanity?
Zhou: They can help humans to accomplish more and more tasks. The main goals of AI development are always dedicated to liberating humans from a heavy workload.
IEEE Transmitter: How much AI research today is applicable versus how much is largely theoretical, in terms of what a robot can actually do?
Zhou: AI is a huge comprehensive subject. Technologies like machine learning, computer vision, voice recognition, and natural language understanding are wildly applied in many cases. Some are still theoretical, such as common-sense knowledge processing.
IEEE Transmitter: Will we see proactive AI in the future?
Zhou: If AI is being used to complete specific tasks, I think proactive AI will soon become reality. However, if we are talking about proactive AI in general, then it still has a long way to go.
I think the biggest challenge for AI is to understand the conversation does not happen in voice recognition technology, but in the capability to understand the complex context in the language of humans.
IEEE Transmitter: In response to AI helping humans with more and more tasks so that we’re not overburdened, what are some examples of the level at which they’d be able to help, not including physical labor?
Zhou: To Liberate humans from heavy workloads – workload isn’t limited to physical labor. For example, radiologists need to review an enormous amount of medical images on a daily basis for disease diagnosis, which is a tough task. AI provides a technical solution to help filtering or bypassing the medical images which are identified as a healthy body, so that physicians will only focus on those images that indicate potential diseases or injuries. This is the way how AI helps to relieve humans from a heavy workload.
Learn more about the impact AI and robotics are having on the world today with the IEEE Interactive Robotics Impact Map.
Written by IEEE on July 22, 2016