Artificial intelligence, robotics, and cybernetics

• Robotics
• Vision
• Machine learning
• Interfaces

Intelligence is the ability to learn, to infer general principles, to communicate in natural language, to plan, and to deal with new or unexpected problems. In at least three of these five tasks, no agreement exists on what it means for an algorithm to do the task well. We still have no standard way to determine which of two algorithms is better at inferring general principles, or communicating in natural language, or extrapolating from known to novel unknown situations. Nevertheless, pseudo-intelligent devices and robots flood the marketplace, and consumers routinely switch from one brand of telephone or robot to another because of small, perceived differences in the intelligence of the two devices.

Researchers in these areas create algorithms capable of learning, generalizing, communicating, planning, and extrapolating. Some study the way algorithms interact with human beings, for example, using robots or avatars as the interaction test bed. Others study the mathematical foundations of the field, trying to find circumstances under which the performance of an algorithm can be guaranteed.

 To specialize in artificial intelligence, robotics, and cybernetics, you’ll want to take: