Communications of the ACM
AI Comes of Age
Artificial intelligence (AI) research has taken great strides in recent years, proliferating and being incorporated into practical applications. Significant AI milestones include the emergence of ubiquitous computing and more powerful computers; software capable of dealing with uncertainty, incompleteness, and anomalies; algorithms that learn and improve over time; and software agents designed to weigh costs and benefits. Among the latest AI innovations is a new generation of software that integrates learning, vision, navigation, manipulation, planning, reasoning, speech, and natural-language processing. Machine learning forms the core of many present-day AI applications, and the availability of vast volumes of information from the Internet and physical sensors has fueled the technology's progress. Carnegie Mellon University professor Carlos Guestrin says that "as the amount of information increases, our ability to make good decisions may actually decrease. Machine learning and AI can help."
Most AI advances have been driven by computer science rather than biology or cognitive science, although Tom Mitchell of Carnegie Mellon University's Machine Learning Department says that new brain studies could enable an unprecedented sharing of information between these disciplines. He observes, for example, that regions of the brain follow pathways predicted by reinforcement learning algorithms used in robots. "AI is actually helping us develop models for understanding what might be happening in our brains," Mitchell says.
From Computerworld
View Full Article
No entries found