Magazine article AI Magazine

Cognitive Orthoses: Toward Human-Centered AI

Magazine article AI Magazine

Cognitive Orthoses: Toward Human-Centered AI

Article excerpt

This introduction focuses on how human-centered computing (HCC) is changing the way that people think about information technology. The AI perspective views this HCC framework as embodying a systems view, in which human thought and action are linked and equally important in terms of analysis, design, and evaluation. This emerging technology provides a new research outlook for AI applications, with new research goals and agendas.


As George Bernard Shaw once observed, being slandered is better than being ignored. So maybe AI researchers should be happy about what's been happening lately. After decades of pundits and philosophers arguing that AI is provably impossible, suddenly that argument has been replaced with the assertion that not only is it possible, but superhuman AI is so inevitable that it is the greatest danger ever faced by the human race. In only about a decade, the conversation has shifted from you can't do it ... to you shouldn't do it! That shift has many parallels in other domains, from vaccination, to flight, to splitting the atom, to gene manipulation. The quest for flight, as we have observed elsewhere, affords a particularly striking parallel. Even in the years the indefatigable Wright brothers were hauling their planes from Ohio to Kitty Hawk to France, ever improving them and demonstrating the improvements, scientific wags--the president of the American National Academy of Sciences no less --were arguing that their quest was impossible. By the 1920s possibility was not an issue, danger was. Books argued passionately that heavier-than-air flight is too dangerous to society, and should be made illegal by international agreement. In capitals around the world, including our own, lobbyists strove to pass laws forbidding attempts at flight. The parallel between AI and artificial flight (AF) is illuminating and suggests that the traditional view of the goal of AI--that is, to create a machine that can successfully imitate human behavior--is wrong.

From the very beginning, attempts at flight sought to imitate the implementation details of birds--the goal seemed to be AB (artificial birds) not AF. In particular, it was taken for granted that flying involved feathers and vigorous flapping. Even in relatively modern times, much scientific debate revolved around exactly how this flapping could be accomplished. The structural similarities and differences of birds and humans had been carefully noted and extensively studied. After all, if you take a skeletal view, people and birds seemed pretty much the same.

Although the Wright brothers were fervent bird watchers, they asked quite different questions, not about flapping, beaks, or feathers, but about lift, stability, thrust, and the physics of turning in air. AI is more abstract than AF but their histories are wonderfully analogous in that both of these strongly held human ambitions were, for a long time, focused on imitating the biological example, and this mistake, in both cases, misdirected these fields. The proper aim of AI is much larger than simply mimicking human behavior. The scientific goal is to provide a computational account of mental ability itself, not merely of human mentality. AI is epistemology, android epistemology. But abstract aims can be pursued apace with concrete applications, and we submit that so far, and in the foreseeable future, most of those applications have been pretty good, largely beneficial for society.

Applied AI does often give deference to the human condition, to human goals and limitations but not necessarily to human mechanisms. A principal goal of applied AI is and should be to create cognitive orthotics that can amplify and extend our cognitive abilities. That is now and near; a computational Golem is not. The articles in this special issue reflect a human-centered vision for applied AI that is less about artificial intelligence and more about amplified intelligence. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.