Quantum Cognition and Other Hogwash

2009.04.18   prev     next

APPARENTLY, William Dembski and like-minded IDists aren’t the only ones infatuated with the idea that human decision must be something “more” than the neural algorithms in our brains, as coded in our DNA. PhysOrg.com has a new article “Quantum Theory May Explain Wishful Thinking”. Choice quotes:

Humans don’t always make the most rational decisions. As studies have shown, even when logic and reasoning point in one direction, sometimes we chose the opposite route, motivated by personal bias or simply “wishful thinking.” This paradoxical human behavior has resisted explanation by classical decision theory for over a decade. But now, scientists have shown that a quantum probability model can provide a simple explanation for human decision-making — and may eventually help explain the success of human cognition overall.

...

In this article, for the first time, we present a fundamentally different, and more powerful, approach to probabilistic models of cognition, based on quantum principles.

...

[The scientists] modeled a game based on the Prisoner’s Dilemma... Here, participants were asked if they wanted to cooperate with or defect from an imaginary partner. Overall, each partner would receive larger pay-outs if they defected, making defecting the rational choice. However, if both partners cooperated, they would each receive a higher pay-out than if both defected.

...

Overall, if our brains use quantum principles, and quantum computation is known to be fundamentally faster than classical computation in computers, then perhaps quantum principles can even help explain the success of human cognition.

I don’t pretend to understand everything that’s talked about in this article (and I doubt most readers — even very intelligent ones — do either), but a few pretty obvious facts need to be emphasized when trying to swallow an article like this one:

1. Describing the act of defection in the Prisoner’s Dilemma as the “rational choice” is seriously bogus. Studies have shown that the best overall strategy in this game is to cooperate on the first move, then copy your opponent’s previous move from that point on. That’s not defecting every time; it might not even involve defecting at all! And if your opponent’s previous decisions are unknown, then the optimum strategy probably involves some simple weighted random choice between defection and cooperation. (Update: Turns out the weights are probably 100% defection and 0% cooperation for this game. But that’s only if you don’t get to know what your opponent has done on previous turns.)

2. Explaining why humans do not do exactly the “rational” thing in this game (even if that is defined as I describe it in paragraph 1 above), is probably a simple mix of these factors:
(a) Human DNA may not be coded to generate a brain that plays a perfect game of Prisoner’s Dilemma without taking a lot of time (months? years? decades?) to extensively analyze the game — time which the test subjects certainly were not given,
(b) Human DNA is subject to deleterious mutations that mildly mess up the programming in many individuals, and
(c) The game is called the “Prisoner’s Dilemma” because the classical description involves prisoners whose decisions about whether to stick to their story, or turn against the other prisoner, will make a huge difference in the quality of the rest of their lives. That’s hardly the case with human test subjects in a psych study. There’s no guarantee that all of them, or even most of them, take the game particularly seriously.

3. So-called “wishful thinking” may just be another way of describing personal preference or desire, which has nothing to do with “rational” decisions. It isn’t “rational” to want or not want any particular thing. Rational decisions are those decisions that maximize the probability of getting whatever it is you want. But that might be any number of things, and none of them are “rational” or “irrational.” They’re just the “givens” of this particular individual. When an individual’s givens strongly conflict with that same individual’s ability to satisfy those givens, then the result will be a profoundly unsatisfied individual, who keeps trying and trying to get that which he likely never will. Such individuals aren’t irrational — they’re just a hazard of this world that, like other hazards, has to be dealt with.

4. There is no model of individual behavior that can’t be described by a clearly defined (i.e. coded) stochastic process. Saying there is such a model is like saying that there may be an entity that can “do the impossible.” There’s no way to absolutely prove that such a statement is false, but it’s so contradictory that it’s practically useless by definition. Wanting to build a school of thought, or a research program, or just a journey of personal discovery, around the idea of non-stochastic, “quantum” behavior may be just another example of the “wishful thinking” (i.e. personal desire) described in the previous paragraph.

Here’s my closing prediction:

Nothing particularly useful will ever come out of attempts to explain human (or animal) behavior as something other than a deterministic/stochastic proc­ess.

See also: The Designed, Cause-and-Effect Brain

Update 2009.07.12 — Another article in the same vein, this time about memristors. Are memristors going to be an important component in future, ultrafast computers? Very possibly. Are memristors analogous to nerve cells’ synapses? Sure, why not. Do memristors contain some special magic that’s needed to create intelligence? No. Intelligence is in the programmed algorithms that use bits of memory, and in the data that is stored in those bits. It’s not in the physical mechanism of the bit itself. Translate the brain algorithms from our DNA to a planet-sized computer made entirely of steam-engine technology — inch-wide metal pipes that store data as pockets of water, and transfer information around with blasts of steam — and though it will run vastly slower than the human brain, it’ll be just as smart.

prev     next