Dibs on why “artificial intelligence” has, is and will fail! (At least if you think “artificial intelligence” means reason like a human being.)
IBM describes the decision making process in humans as four steps:
- Observe
- Interpret and draw hypotheses
- Evaluate which hypotheses is right or wrong
- Decide based on the evaluation
Most of us learned those four steps or variations on them as part of research paper writing or introductions to science. And we have heard them repeated in a variety of contexts.
However, we also know that model of human “reasoning” is a fantasy. Most if not all of us claim to follow it but the truth about the vast majority of decision making has little to do with those four steps.
That’s not just a “blog opinion” but one that has been substantiated by years of research. Look at any chapter in Thinking, Fast and Slow by Daniel Kahneman and tell me how Watson’s four step process is a better explanation than the one you will find there.
One of my favorite examples was the impact of meal times on parole decisions in Israel. Shai Danzinger, Jonathan Levav, and Liora Avnaim-Pesso, “Extraneous Factors in Judicial Decisions,” PNAS 108 (2011): 6889-92.
Abstract from Danzinger:
Are judicial rulings based solely on laws and facts? Legal formalism holds that judges apply legal reasons to the facts of a case in a rational, mechanical, and deliberative manner. In contrast, legal realists argue that the rational application of legal reasons does not sufficiently explain the decisions of judges and that psychological, political, and social factors influence judicial rulings. We test the common caricature of realism that justice is “what the judge ate for breakfast” in sequential parole decisions made by experienced judges. We record the judges’ two daily food breaks, which result in segmenting the deliberations of the day into three distinct “decision sessions.” We find that the percentage of favorable rulings drops gradually from ≈65% to nearly zero within each decision session and returns abruptly to ≈65% after a break. Our findings suggest that judicial rulings can be swayed by extraneous variables that should have no bearing on legal decisions.
If yes on parole applications starts at 65% right after breakfast or lunch and dwindles to zero, I know when I want my case heard.
That is just one example from hundreds in Kahneman.
Watson lacks the irrationality necessary to “reason like a human being.”
(Note that Watson is only given simple questions. No questions about policy choices in long simmering conflicts. We save those for human beings.)