From Eurekalert, a press release titled, “Complex decision? Don’t sleep on it”:
Neither snap judgements nor sleeping on a problem are any better than conscious thinking for making complex decisions, according to new research.
The finding debunks a controversial 2006 research result asserting that unconscious thought is superior for complex decisions, such as buying a house or car. If anything, the new study suggests that conscious thought leads to better choices.
Since its publication two years ago by a Dutch research team in the journal Science, the earlier finding had been used to encourage decision-makers to make “snap” decisions (for example, in the best-selling book Blink, by Malcolm Gladwell) or to leave complex choices to the powers of unconscious thought (“Sleep on it”, Dijksterhuis et al., Science, 2006).
At stake in these conscious/unconscious thought experiments (literally) is a wider philosophical argument about the value of intuition and hunches. We want to think that hunches produce better decisions, and have been taught since we were children that this is an intelligent way to approach reality (“Use the Force, Luke”.) However, it ain’t so. Though hunches may be useful for simple decisions, like when to swing a bat to hit a ball, conscious thought appears to be superior for complex decisions, the ones that really matter.
It appears that the mysteriousness of unconscious thought may be part of its appeal. However, I find that conscious thought can be as mysterious as unconscious thought. Underlying every conscious thought is a bedrock of unconscious beliefs and assumptions. Only through deliberate questioning can we methodically dig up these beliefs and question them for accuracy and relevance. Without regular housekeeping, things can get pretty messy down there. The great project of analyzing our beliefs with conscious thought is far more interesting than the plug-and-play autonomicity and quick fix of unconscious thought.
Some arguments for the infeasibility of AI rest on the supposed mysteriousness and power of unconscious thought. But as I mention, conscious thoughts rest on unconscious ones, so this mysteriousness and power are still retained in consciousness. All that aside, cognition is way less mysterious than it was a few decades ago, and now we know a tremendous amount about the mind. It’s only a matter of time before its structure becomes understood, just like our place in the cosmos, interactions between chemicals, the behavior of electromagnetic fields, and thousands of other phenomena that were once baffling but are now taught in High School.
Of course, investigating the structure of thought in greater detail and coming to understand it may frustrate people like Douglas Hofstadter, who would lose respect for humanity if we come to learn too much about ourselves too soon. According to Hofstadter, reaching the goal of AI in a few decades would make him “fear that our minds and souls were not deep”.
Such spiritualistic language in reference to the human mind only discourages level-headed research and objective question-asking.