Artificial Intelligence and the Future of EmploymentUsing Artificial Intelligence (AI), machines are now matching or outperforming humans in a number of tasks by which humans have traditionally earned a living. (Note, it’s not just factory and driving jobs which are threatened. AI is currently matching/outperforming professionals various ways in the fields such as banking, law, and medicine.) Assuming that AI and automation in general will eventually reach a point where many, most, or even all traditional jobs are better done by machines than by humans, what might happen? Consider the two answers below.Jack: We could have a Robot Utopia if AI began to outperform us in most of these tasks. Drudgery would be eliminated. Diseases would be more effectively treated, more quickly eradicated. In general, most if not all of the problems humans have always had to work long hours over many years trying (sometimes futilely) to address/solve would be getting better addressed and solved by machines that are simply better at it. And we would be free to pursue our dreams, living lives that are more enriched, more secure, more adventurous, more full of whatever things we actually like to do. Let’s work toward this! And if you think I’m blindly optimistic, just consider how advancements in society have already addressed these problems and created more free time than past humans would dare to dream of.Jill: Robot Dystopia is more like it. Without the extrinsic value we get from our careers and without the sense of making important contributions to society and relying on each other, we might be able to have some fun times, but ultimately society would fall apart. At some point the vacation must end. When things get too easy, life becomes drab and meaningless. I’m afraid we’d see more depression, ennui, and amorality. We should do whatever we can to avoid the human obsolescence which would accompany what Jack calls a Robot Utopia. In no more than 300 words, state whether you agree with Jack or Jill, or whether you have an answer of your own which is significantly different. Whichever of these options you pick, explain your reasons.Solving EthicsAs the world becomes more automated, actions that had always been limited to humans are now being taken by machines, and with serious consequences. In just the last couple of years, we’ve seen the stock market disrupted in microseconds by algorithms, racial biases emerging in algorithms, one pedestrian and some occupants killed by automated vehicles, and now over 300 people killed when, on two separate occasions, pilots of Boeing 737 Max 8s “struggled to control [their] aircraft as the automated MCAS system repeatedly pushed the plane’s nose down.” (I’ve included the links for support and your perusal if you’re interested, but you are not required or expected to read through all these articles.)Until only very recently, if some individual X exhibited a racial bias in finance or medicine, or if X drove a car or took over a plane so as to seriously jeopardize the lives of others, this X was a human, and that human could be blamed, criticized, and punished. Humans can be deterred from such actions by these kinds of consequences, and it makes sense to exact justice on them when they are not so deterred. But what should we do now that X can be an algorithm or machine? How can we ensure that our automated, intelligent, and seriously consequential artificial systems behave as morally as possible?In no more than 400 words, please solve ethics for our increasingly automated future. Kidding! But try to be as specific and thoughtful as you can about how we could best face this challenge. For some guidance: Is significant regulation required, and if so who would be the regulators? Should governments form treaties or work it out independently? How should an increasingly automated society manage and reflect our varied and often conflicting values? Is this a problem we should solve democratically, with panels of experts, or in some other way?If you see connections to concepts we’ve encountered in the past (moral hazard, precautionary principle vs. liberty, the Original Position–to name some) it might not hurt to acknowledge them.
