Digital technologies for learning, health, politics, and commerce have enriched the world. Digital heroes like Sir Tim Berners-Lee, Batya Friedman, Alan Kay, JCR Licklider, and Joe Weizenbaum have blazed trails. Yet there is trouble. We depend upon software that nobody totally understands. We are vulnerable to cyberterrorism. Privacy is overrun by surveillance capitalism.Totalitarian control advances. Daily internet news matching our beliefs makes it hard to tell true from false. People are addicted to devices. Jobs disappear without social safety nets. Digital leviathans threaten to control all commerce. Our values are threatened.
There are risks of premature use of AI in domains such as health, criminal justice, senior care, and warfare. Much current AI is unreliable, without common sense, deceptive in hiding that it’s an algorithm, unable to explain decisions, unjust, neither accountable nor responsible, and untrustworthy.
A forecasting model is a prediction of how the world will evolve, of what will happen in the future with respect to some phenomena (such as the motion of objects, the financial health of a business) or the spread of an epidemic.
For most of human history, dyads and groups were only able to work and play together if they were collocated. All of this changed in the 19th century, when the first remote collaboration and entertainment technologies — the telegraph, the telephone, and the radio — were developed and widely commercialized. These were joined in the 20th century by television. By the middle part of the century, medical images were being transmitted over phone lines; soon thereafter, 2-way television was being used for remote medical consultations.
In this column, in my textbook, and in a speech “What Society Must Require from AI” I am currently giving around the world, I document some of the hype, exaggerated claims, and unrealistic predictions that workers in the field of artificial intelligence (AI) have been making for over 50 years. Here are some examples. Herb Simon, an AI pioneer at Carnegie-Mellon University (CMU), who later won a Novel Prize in Economics, predicted in 1958 that a program would be the world’s best champion by 1967. Marvin Minsky of MIT, and Ray Kurzweil, both AI pioneers, made absurd predictions (in 1967 and 2005) that AI would achieve general human intelligence by 1980 and by 2045. John Anderson, discussed below, made the absurd prediction in 1985 that it was already feasible to build computer systems “as effective as intelligent human tutors”. IBM has recently made numerous false claims about the effectiveness of its Watson technology for domains as diverse as customer support, tax filing, and oncology.
Every Computer Science student should get significant exposure to the social, political, legal, and ethical issues raised by the accelerating progress in the development and use of digital technologies.
The standard approach is to offer one undergraduate course, typically called Computers and Society or Computer Ethics. I have done this during the current term at Columbia University, using my new textbook, Computers and Society: Modern Perspectives (OUP, 2019). We meet twice a week for 75 minutes. In class, I present key topics covered in the book, and welcome a number of guest speakers who present their own experiences and points of view. Every class is interactive, as I try to get the students to express their own ideas. There have been four assignments: a policy brief, a book report, a debate, and a research paper. Such courses are typically not required by major research universities, which is a mistake, but they are often required by liberal arts colleges.
Many issues discussed in Computers and Society: Modern Perspectives suggest a need for legal remedies, such as the case of monopoly power in digital technology industries. Other issues raise ethical quandaries, such as the cases of employees of such firms who find actions of their employers immoral. In almost all cases, such as technology addiction, fake news, and unjust algorithms, wise legal actions and informed moral choices depend upon having good information about what, how, and why things are happening. This requires research. In an excerpt from his excellent recent book The New ABCs of Research: Achieving Breakthrough Collaborations, published by Oxford University Press, Emeritus Prof. Ben Shneiderman suggests that what is needed is applied research illuminating context and situations coupled with basic research illuminating causes.
My textbook — Computers and Society: Modern Perspectives — may be used in a variety of courses and contexts, but is intended primarily for use by Computer Science (CS) Departments, as they attempt to educate and train tomorrow’s software professionals, managers, and IT leaders. If we want to monitor how well departments are doing this job, we should ask is if they are sensitizing their students to the ethical responsibilities of the profession. It is useful to contrast the attitudes and performance of CS Departments, typically situated in science faculties, with departments in Faculties of Engineering.
Concern over ethics in Engineering began after several major disasters late in the 19th century and early in the 20th century, notably several bridge failures and the Boston molasses disaster, in which a flood or molasses wreaked havoc on nearby building and train systems. There already had been created professional societies such as the American Society of Civil Engineers and the American Institute of Electrical Engineers. These societies then moved quickly to introduce Codes of Ethics and requirements for licensing and accreditation, which ultimately caused university departments and faculties to include some learning about and practice with ethical concerns as part of their curricula. A later development was the creation in 1954 by the National Society of Professional Engineers of a Board of Ethical Review.