When people are compared to computers, one means either the excellence of the human mind over the primitive "mechanical" thinking, or, conversely, the poor efficiency of humans as compared to computers in many "intellectual" tasks. And one can hardly find a sober reflection of the very distinction of the two mental styles. Does the computer really differ from humans in their way of thinking? If it does, what is the difference? Is this difference a permanent feature or it will develop into something else in a while?
Machines are usually praised for their consistency, the ability to follow the pre-set rules of the game. In humans, such rigidity is almost impossible, we depend on our moods; we cannot concentrate on a single task, being engaged in many interfering activities. But the latter habit makes us more resistant to the pressure of circumstances, and human won't just "hang" in the heavy flood of service requests, as computers often do.
From the formal viewpoint, humans are very chaotic and uncertain. They do not even conform to a stochastic picture. The human history is replete with the attempts to establish some universal order, and the dramatic stories of its inevitable violation. Computers are more likely to be the law-abiding citizens of the perfect state; most humans would take such a state the worst ever dictatorship.
The principal difference between the two styles is in its primary direction. Computers are inner-bound, their "motivation" comes from inside and hence it can only follow the already existing models of behavior. Humans are essentially outer-bound; they have to follow the variations of the objective reality, obeying rather the laws of nature than any formal prescriptions. These two attitudes complement each other, and an outer-bound behavior is often necessary to initiate an inner-bound activity, while an inner-bound functioning can sometimes pose questions to an outer-bound mind.
The distinction in the logical orientation does not directly depend on the formal complexity of the problem. Very complicated tasks can be efficiently solved by machines, provided there is no need in additional considerations of anything beyond the current scope. On the contrary, the necessity to adapt to the "contradictory" demands from the outside may hang a computer in a relatively simple case.
We already know that the elements of the inner-bound logic can saturate the human activity in the conditions of relative historical stability. One could predict the appearance of the outer-bound logic in computers on a higher level of development, with the computer systems transforming into a kind of "computer society". For this new society, humans will serve as an inner-bound system producing standardized reactions to complete the objective motivation of conscious computers.