Viewpoints

About This Technology

In some cases, when a task requires intelligence, artificial systems can do the job. Artificial speech recognition as well as face- and handwriting-recognition software rely on techniques that emerged from AI research. Other everyday applications of AI include a virtual pet that can learn from its experience and artificial characters in video games that change their skill level to adjust to a user's abilities. AI also enables autonomous vehicles such as those that run in DARPA Grand Challenges and European Land Robot Trials. Developers aim to solve increasingly difficult challenges—such as analyzing security-camera images and providing a humanlike agent that delivers natural-sounding speech answers to natural-language queries. Some software development aims to enable humanlike abilities—to translate speech and to guide robots to walk on unpredictable terrain and manipulate objects. Other AI development efforts aim at applying intelligence toward goals that would be difficult or impossible for people to achieve: Credit-card–authorization systems make fast decisions about whether to allow a purchase at a retail store, automated financial-trading systems arbitrage multiple stocks that a human could not keep track of, and AI software advises doctors in diagnosis, relying on knowledge bases that are more comprehensive than a person can reliably memorize. AI may even be evolving into an uncanny capability that does not emulate human intelligence but instead seems to have a mind of its own—as in the case of chess-playing computers, which have distinctly different styles of play from those of human chess players.

Researchers and developers take various approaches toward endowing artificial systems with intelligent attributes. The most mature approaches to AI simulate elements of human intelligence such as reasoning, learning, planning, and movement. Much research revolves around artificial reasoning: Software can prove or disprove mathematical propositions from axioms, and very similar software can make inferences about everyday matters from commonsense propositions in a rigorous formal language. A second family of approaches rests on the idea that intelligence may emerge from systems that emulate biological processes—systems such as artificial neural networks and genetic algorithms. Such methods appear in data mining, antenna design, simulation of crowd scenes in movies, and industrial-operations research. A third family of approaches that has yielded many practical AI-like applications arose from outside the mainstream of AI research, from the field of applied mathematics—especially statistics, probability, and optimization. And R&D in a field that's closely related to AI—the Semantic Web (whose advocates insist that it is "not AI")—aims to encode Web pages in machine-readable form, such that machines can determine the relationships among the concepts that are also in natural-language (human-readable) form. A Semantic Web inference engine can then draw conclusions, perhaps along the lines of "the spider cannot be a poisonous black widow because the university Web site indicates that all black widow spiders have red marks, and the student's Web site indicates that the spider in question is black and gray." Finally, a major direction for AI relies on combinations of all approaches or ad hoc approaches—anything that works—to perform tasks that people consider to demand intelligence.

Current progress in AI promises to enhance transportation and robotics, office productivity and collaboration, physical security and information security, transactions and logistics, entertainment and advertising delivery, and social interaction. Cars will rapidly respond to dangerous situations to avoid collision. Software will coordinate movements of service robots to perform useful tasks—perhaps gracefully. Personal-information agents will assist with scheduling and organizing meetings. Image-recognition systems will retrieve digital photographs, automate content production, alert people when security cameras capture events of interest, and provide context-sensitive information, entertainment, and advertising. Yet AI research also produces disappointments as developers encounter seemingly intractable problems such as language translation and commonsense reasoning. Visionaries and optimists may be disappointed with near-term projects to identify terrorists before they commit terrorist acts and to build robots that can provide safe and effective care of elderly and frail persons. But a good number of AI challenges are intermediate between intractable and feasible. Decision makers can benefit from monitoring advanced developments to reduce uncertainty about this large portfolio of intermediate-level problems. For example, different outcomes will result depending on the success of AI developers' efforts to manage risks in financial portfolios and credit decisions, generate content and advertising that is truly relevant to individual users at specific times and places, accept training from people who are not technology experts, and train robots to navigate through doors and handle food and drinks. Different outcomes could result depending on whether users will find software agents to be trustworthy when executing transactions, taking control of cars, and monitoring security cameras; whether people will consider that agents provide believable responses to conversational queries; and whether interacting with software agents in virtual environments is worthwhile.