Over the past few years, Communications of the ACM has published a number of short articles related to various aspects of artificial intelligence (AI). This particular article concerns the effect that voice and facial recognition is having on society today. Twenty-five years ago, voice recognition software was still in its relative infancy, often being trained by its user’s voice one character or word at a time--and for use only when that voice didn’t have a cold! The author reports that, in the first quarter of 2018, Google and Amazon sold 5.7 million voice-assisted devices. There are few people alive today who are not affected by such machines.
However, despite the many positive aspects that voice and facial recognition systems have had in some parts of society, there is also a dark side: totalitarian governments (among others) are using these systems to control the populace; privacy concerns abound in cities, which have cameras monitoring every street corner. The author also reports that the Electronic Frontier Foundation published a whitepaper in which they claim that minorities are being “misidentified ... at higher rates than whites” [1].
The final section of this article, “No Easy Answers,” discusses how difficult the road ahead is in combating these issues. There are no references, but plenty of quotes and five papers listed under “Further Reading.”
This article can be useful to information technology (IT) educators, students, and professionals in discussing some of the ethical issues that AI systems can raise.