Artificial Intelligence (AI) Nightmare Scenarios: Summoning the Demons?
Some may think of Artificial Intelligence (AI) in terms of Boston Dynamics’ Robot dog, Spot or of software that can defeat the very best human player in Go or Jeopardy. This course is about a much more sinister and serious aspect of AI, the various “nightmare scenarios” that are appearing regularly in the technical literature and sometimes in the daily news. These scenarios take many forms, possibly the most familiar being the use of “killer robots” and other military applications where the battlefield, and military strategy are dominated by AI applications. Thought leaders as varied as physicist Stephen Hawking and entrepreneur Elon Musk have vehemently warned about the potential dangers of this situation, which Musk referred to as “summoning the demons”.
There are many other worrisome scenarios, like the prospect of AI causing massive global job dislocation and unemployment as more and more jobs are automated. The benefits and dangers of super intelligent surveillance systems, encouraged in some nations like China, but legally restrained in some large metropolitan cities in the free world, like San Francisco, are often in the news. MIT researchers have found that some of the highly sophisticated facial recognition platforms are ethnically biased, an interesting and important research issue. And the increasing cleverness of “deep fake” technology makes it possible for a seemingly authentic video from a world leader to be completely fabricated by an enemy. And there are increasing risks associated with AI-generated decision-making algorithms in fields from medicine to insurance to the stock market. Funds run by these algorithms currently account for 60% of stock trading activity, according to The Economist.
In this lecture we will examine the ideas of the best-known thought leaders on these topics, and also have plenty of discussion among ourselves about this crucial topic.
A/V requirements: LCD Projector, Screen, Laptop