The end of uncertainty and a new beginning.
The end of uncertainty and a new beginning.Big data seems unstoppable (see this illuminating essay in Spiegel Online International). As data – and that means data about YOU – grow exponentially, your behavior is becoming ever fully mapped. What you buy, what you consume, where you travel, it’s being stored in the ocean of zetabytes. Present-day prophets proclaim “the end of theory” and “the end of chance.” They predict that Laplace’s fantasy will become reality. Laplace’s demon (also known as Omniscient Jones in the Anglosphere) will reside in the data tanks and their algorithms.
The end of chance (I have commented on the end of theory here) – well cry me a river. Where are the objections from the Humeans or the storm troopers of quantum physics? This would be a good time for followers of Hume to point out that induction, that is, predicting the future from the past, will remain logically indefensible even if in practice predictions rarely fail. Lovers of the quantum will claim that there will always be irreducible unpredictability, as in the atom, so beyond.
As the efficiency of big data and their (not its) mining continue to grow exponentially, the mapping of your past behavior takes the mystery out of your future “choices.” I am putting “choices” in quotation marks because the acolytes of the Church of Free Will will have some explaining to do. When you walk into a department store and you are greeted by a robot who gives you a coupon for the very item you are about to “intend” to buy, you realize with a sinking feeling that the robot knows more about you than you do. The robot, with its algorithms and access to big data that suffer no memory loss, has the advantage.
Enter trust. Our friends in the financial industry (bankers, insurers, loan sharks) will know better than you do whether you are not only able but also willing to repay your debts. Id est, they have no more need for trust. They know.
If this head-spinning rush toward a full victory of determinism were to be democratic, you too would be equipped with a robot that can run algorithms on big data. You would be able to predict the behavior of merchants, loan sharks, and police officers. There is a small voice of hope and it says that once technology is unleashed, we will all have it.
All too often, we hate paradoxes because of their potential to induce headaches and because of their disdain for logical coherence. But here’s one that gives me cheer. If big data takes the uncertainty out of human behavior, and if access to it is uniform, then its victory will nullify itself. If I can predict what you will do and you can predict what I will do, one of us can go to the second level and deliberately (by “choice”) do the opposite. The other, knowing this, can counter. Et cetera, ad infinitum. The result is that both our behavior returns to unpredictability.
The end of uncertainty? Guess again!
For an entertaining treatment of the idea that full mutual predictability will remain elusive, read The Princess Bride; for a more formal treatment, see Sir Karl Popper's ideas on The Oedipus Effect (not Complex) in The Poverty of Historicism. For the idea that we humans have some room to limit how predictable we are to others, see earlier posts here and there.
No substance. A Bergamascan lawyer told me with confidence that psychology has no substance. Before I could respond with an eloquent slap on his orbitofrontal cortex, he added that psychology has function. I settled for that. -- I should have said, 'Yes, it's all in your head.'
Years ago, a lawyer told me that psychology has no utility. What’s wrong with lawyers?
Expectations. They say you should expect the unexpected. I thought about it, but couldn't figure it out.
A third-person ultimatum game.
A third-person ultimatum game.In the standard ultimatum game, one person (the proposer) makes an offer about how to divide a sum of money (e.g., 10 ducats) between himself and another person (the responder). If the responder accepts, the money is divided as proposed. If he rejects the offer, neither person gets anything. Most proposers offer a 50:50 split and most responders accept. The average offer is somewhat lower. Offers below 1/3 are progressively likely to be rejected. Such rejection is irrational from a standard economic point of view, which holds that something is better than nothing. Social preference theories try to explain this result by assuming that people care about equality and hence resent a biased proposal. By refusing a biased offer, responders punish the proposer more than they punish themselves. Doing this, they may hope to reform the proposer's selfish attitude and send a message that they, the responders, are not to be trifled with. They manage, in other words, their reputation as partners in social exchanges.
The standard two-person ultimatum game confounds the responder’s preference for an equal outcome with his resentment for the proposer’s nerve to claim the lion’s share. This confound can be removed by introducing a third person. The third person, let’s call him MB (Mr. Moneybags) proposes to split the 10 ducats by offering 9 to Mr. A and 1 to Mr. B. The stipulation is that the split will be executed as soon as both agree. Otherwise, Mr. Moneybags keeps the stash. Now, Mr. A is likely to consent, although he is not certain to do so. If the social preference for inequality aversion is a powerful force, he will reject the offer in order to send a signal to MB. Mr. B is more likely to reject the offer. If he does, it is presumably because he objects to the grave inequality. His rejection brings a punishment of Mr. A as a byproduct, but it is not a moral punishment because Mr. A has done nothing wrong. It would be an interesting experiment to run the standard and the third-person ultimatum game side by side to see how much more inequality responders will tolerate in the latter. If someone knows of a study in which this was done, please leave a comment.
Güth, Schmidt, & Sutter (2007) describe a three-person bargaining game, in which the first person proposes a three-way split between himself, the responder, and a dummy player. Here, the proposer remains in interested party. As in the two-person game, the modal outcome is the proposal and the acceptance of an equal split.
Illusion of control, sort of.
Illusion of control, sort of.When I go to my desk at the University of Bergamo, I need to get past a gate. Every morning, I press a button of an intercom. Usually, a voice comes on that says “Prego.” I then offer a random sample of 1 of the following gambits: “Aprire la porta per favore per il professore.” “Professore Krueger,” or just “Professore.” Or “Please open the gate.” Once I toyed with the idea of saying “Professore Mickey Mouse, per favore.” The intercom lady usually obliges and opens the gate. One time she was away from the mic, though, or not in the mood. That day, I had to take the long road. My broader point is that often, there is a loose correlation or no correlation at all between what we do and what happens. Yet we press on, acting as if our actions had causal power. We know this irrationality as the illusion of control. At the limit, imagine a situation in which what you do is perfectly unrelated to what happens. The variability of your behavior only has esthetic value, but no utility. Once you act, chance reigns. But all is not of naught. Perhaps it matters that you do something in order for the porta to aprire. It’s like buying a lottery ticket in order to give chance a chance. If you don’t play, you can’t get lucky. Certainly, this is not a particularly novel point, but one worth making repeatedly at unpredictable intervals. So get out there and do something.
Babylon. Growing up in the Humboldtian educational system, I was forced to endure 5.5 years of Latin. It was what one did. A friend of mine asked a teacher why we had to study Latin. The teacher said "Because we must." That summed it up. The Prussian culture of obedience in its pure form. I always felt that learning a live romance language would have had greater utility. If you know Italian, it can help you study French, no? Why begin with the dead root? Finally, 40 years later, I am harvesting late-ripening fruit. Teaching psychology in Bergamo in English presents a rare indulgence. I can refer my students to their Roman ancestors and what they would have said. The last mot de jour was ceteris paribus. I enjoyed how my Italian students pronounced it (cheterisse paribusse). Naively, I thought that Italian, which in good Humboldt (Wilhelm, that is, not Alexander) tradition I regard as a form of Vulgar Latin would have a recognizable version of this phrase. Apparently, it does not. I went to the Babylon translation machine by way of the google company and asked for an English - to - Italian translation (as self-respecting non-Humboldtians, they don't do Latin). They say that ceteris paribus is "a parità di tutte le altre circostanze." I hate to say it, but a feeling came over me that made me miss good old Latin.