Ever wonder if sometimes the computer has a mind of it’s own? I personally never gave that much thought. Others have claimed that when I walk into a room where they show me a broken computer, it just all of the sudden behaves. I still say those were just coincidences.
Every once in a while a solid story appears that makes you wonder. Separating Programming Sheep from Non-Programming Goats is an article that mentions unsubstantiated claims that says people who program well have a good mental model of programming. People who make good programs and work well with electronic technology also do not see personalities in machines. That is what a related study says.
The core tenet of what makes a computer a computer goes back to what Alan Kay said about the machines as recently as March 2019. His exact words were: A computer is a process that deals with represented descriptions. Elsewhere he refers to Computer Science as the science of processes. His words contains many layers to unpack which I will not do here but the main interest among programmers is the definition and implementation of processes.
A process is itself a parameter (constraint) on the machine that expresses it. A computer can be considered a relatively unbounded mechanism to enact bounded operations on data. That the machine in its entirety in terms of every circuit, transistor, and actual voltage variation at each level from moment to moment is seldom understood by a single person in a single thought is common enough. In ordinary practice, selecting from the unbounded set of bounded operations for a given process often proceeds from accepted computational models and code practice norms.
Programmers use layers of structured logic to define process that work effectively well as computer programs. Usually, this means there is no magic involved and no mystery in how the computer works. However, are there situations in which intuition, experiential observations, and instinct complement logic in order to better understand and define computable processes?
A post on Quora present stories from established software developers in which they saw unexplainable situations involving computers. At least not explainable in a way in which certain things could be repeated in a typically logical fashion. What they reveal is that we can say things sometimes get so complex that you cannot explain them the same way you would simpler forms and principles. As an example, Windows 10 has superficial similarities to Windows 95 but earlier simplicities give way to increased complexity over time. We can fully explain MS-DOS running on a 486 system. Yet, a complete explanation of the Tesla Autopilot System eludes even its creators.
Where does the disconnect in understanding originate? Perhaps the answer surrounds us after all. The ingredients we use to make physical computers upon which software relies comes from nature itself. Though we can explain much of nature, we cannot explain all of it let alone understand it in full.
Quora question … Is true that some code works, but no one knows why? A great discussion from established software developers about computer phenomena that was not entirely explainable. The best solution in some cases was singularly intuitive. Also, a very short answer to the question … What do you find obvious but science or society hasn’t accepted yet? in which the response was the universe itself is probably not logical but illogical. Then there is this gem … Can technology become too complex that the civilization that created it can no longer understand it? How can humans delay or prevent such an outcome?