∫ Philosophy of computer science: the dawn of thinking machines

Like all other things and unlike any other things, the origin of computers started with a question, but it's a unique question that had been troubling the greatest of minds since the origin of human philosophy.

In order to appreciate the difficulty of how the question developed, let's first have a look at some others. Among the other questions, there were people who questioned our origin, and there are people who wondered what happens after we die. Do the consciousness really exists as a separate entity from the body? Or we all are what we have in life only? We already knew more or less satisfying answers to some questions. Like the question of our origin, the first few attempts to answer it had let to religions, but it is not until Charles Darwin proposed and verified the evolution theory that we have the answer that follows the scientific paradigm. The same attempt to answer the trace of soul after death did not go as smooth, and they remain the problems which left unanswered if not unable to answer, largely due to the fact that we have no means to verify the answers provided.

But there is another question that had been answered to some degree but incomplete, not because we couldn't verify. But we didn't know whether we are qualified to verify it at first.

What is this question that although human mind can conceive but we were not sure how to verify? Unfortunately it is the question of the scope thought itself: how far can we think? We could hypothesize many things, come up with great theories to explain the universe, but when it came to the scope of thinking, how could we be sure that the answers provided by the same apparatus that poses the question are trustworthy?

It turned out that we can! but not through and through, thanks to the contribution of Kurt Gödel, Alonzo Church, and Alan TuringDyson, G. (2012). Turing's cathedral: the origins of the digital universe. Pantheon.. They had shown that, which we will see, within the scope of thinking, we can "verify" many things even the thinking itself, but not all of it. Over the past fifty years the challenge had gradually shifted, from what can we verify with thoughts, to the original question of why do we think, and how do we think, and even further beyond to how do we improve thinking, or how do we improve the process of improving thinking?

It is all these attempts, to answer the question of our mind in both why and how, that had led us to another scientific revolution, which enabled the invention of computers when they were needed during World War II to decrypt the Enigmas. After the war, this revolution not only persisted but flourished. Led by the most brilliant mind of era, John Von Neumann, a human computer, we used computers to calculate necessary conditions for nuclear chain reactions, filter noise from radar to improve radio detection, and even brought humanity to the moon! Computers had continued to transform our lives ever since. Nowadays, from waking up in the morning to going back home from work everyday, there are computers in almost every aspect of modern life. The utility of modern computation engines is anything but waned.

And now we shall go back to the very first question that pushes the society since the beginning.

∂ How far can we think?

Knowing the answer to this question will allow us to understand the limits of our imagination, know what can and can't be verified, before we could ever hope to move beyond.

For the comparison, there are questions about thought which are considerably simple. For example, what is thinking? Thinking is not hard to define. Like when we are asked to explain what is sweet. We could do it by comparing to other tastes. Likewise thinking is a series of transforming information in our head step by step. We could compare it to perceiving or acting to appreciate the difference. But when it comes to the question of how further can we think? Just like when we test our eyesight with a snellen chart, we will require some assistance from an external tool that could express something more complicated that our thoughts.

Fortunate for us, not only did the tool exist at the time the question was answered, but ever prospered since the first time it was constructed by Babylonians as the basis for their calendar system. That tool was mathematics.

Even if mathematics could be more expressive than our thoughts, to acquire new knowledge in mathematics, we need to prove. Proving is a process of verification, by nothing but our mind. If we can prove something in math, that's just the same as our thought could verify it. Therefore the limit of mathematical proofs would spell out the limit of our thoughts.

David Hilbert was a renowned mathematician of the early twentieth century. Given his success in measure theory, analysis, and topology, he was ambitious to formalize all the mathematical proofs to the point that they could be attained without paradoxesEllenberg, J. (2015). How not to be wrong: The power of mathematical thinking. Penguin.. Another way of saying that everything can be proved. This ambition had been thwarted first by Kurt Gödel, then by Alonzo Church, but it was an english mathematician who scored the finale…

Alan Turing was wondering about the same thing. Though as a son of a mere civil service clerk, Turing had displayed a prominent aptitude in schools since his early ages. Being a true servant of mathematics he could have derived calculus, if not for the era he was born into where it had already been discovered. At the time, mathematics had advanced to the point that people started to wonder, was there anything that we couldn't verify. And Turing showed them that there was and there would always be. Through the device that he conceived he could construct a scenario where his device would fail to give us the correct answer. What's interesting from our point of view is that, this device, as we now refer to as the Turing machineMartin, J. C. (1991). Introduction to Languages and the Theory of Computation (Vol. 4). New York: McGraw-Hill.Chicago, is an imitation of human thinking process. Then latter Turing would turn out to be known as a founding father of artificial intelligence, but not before he shattered the poor Hilbert's and many other mathematicians' dream.

After Turing, the dream of thinking beyond what we can think, has come to an end. But there's a light to it. Within the proven scope of thinking, there's a room left for thinking about thinking itself! Because a Turing machine can simulate another Turing machine! If this means anything, it would seem we could conceive a way to improve thinking! But how to do it in the most efficient way?

To answer this question, we must first understand why nature has bestowed us the ability to think in the first place. What are the problems in our lives that nature was so desperate to enhance our brain to such an extent?

∂ Just why we think?

Every ardent reader must have came across a quote that goes,

It was first uttered by a french philosopher, Rene Descarte, in his treatise Discourse on the Method, as an attempt to understand thoughtDescartes, R. (2006). A Discourse on the Method. OUP Oxford.. Though Descarte simply conferred its importance to our existence, not so much an explanation after all.

In order to rectify Descartes' answer, we must go beyond what he knew. Thanks to the new insights in physics, we now understand the concept of entropy, which was introduced after Descartes time, by a scientist of the same nationality of his, Lazare Carnot, and contemporarily together with Darwin's theory of evolution.

For everything within the universe, entropy is ever increasing by the rule we now know as the second law of thermodynamics. Thinking, as a mechanism in our brain, which in turn a product of evolution that follows the same law since the beginning of the universe, must therefore exists by the very same law. But natural selection does not always result in thoughts; as there exists classes of living things that we are pretty sure that they can't think, like those organisms of simple cells. Therefore you shouldn't define the existence of something by its ability to think!

Still animal actions are quite definitive. If we don't think about it, what we actually want to do in our lives, guided by the innate sensation alone, is to consume more and reproduce more so that we can consume even more! The consumption and dispersion of high energy content cause the entropy to increase, all of which are guided by the pursuit of pleasure and the avoidance of displeasure. Evolution has managed to route a shortcut for us to increase the entropy.

But then why thinking? Since the nature allows thinking that means having it should somehow promote its trend to create more lives and therefore further accelerate the increasing of the entropy.

Roughly speaking, it is because of thinking that increases the chance of our existence. Our ancestors thought, hence we exist! We think, therefore we are! Not of the meaning by Descartes, but by the means of survival of the fittest.

But what's the thing that we do to find pleasure in life? Emphasize on the word "find". That's it, search! We search for pleasure. We search for more of it. And we search for ways to acquire more of it!

∂ Think to search better

Ever since homosapien had crossed the threshold of evolution and became sentient, all the problems that mankind ever faced are search of some kindNewell, A., & Simon, H. A. (1972). Human problem solving (Vol. 104, №9). Englewood Cliffs, NJ: Prentice-Hall., whether it is for something or somehow. Some problem is easy, especially when there are not so many choices to try. Some problem is harder, because we have to invest in a search before we could get anything.

If we have an infinite amount of resources, searching is trivial, we could simply try one thing then another, then another, and so on until we find what we seek. Oh, well, but infinite is just a mathematical trick. Everything in our observable universe is finite. Even the time itself has the beginning and the end. We could only hope to do search with the limited resources.

Managing resources efficiently when searching is a kind of intelligence. In fact, it defines intelligence! This is not the first attempt though. Intelligence has been defined by many people on countless occasionsIntelligence. (2020, January 24). In Wikipedia. Retrieved from https://en.wikipedia.org/wiki/Intelligence. Howard Gardner related it to problem solving. Robert Sternburg and William Salter thought of it as a goal-oriented adaptive behavior. Shane Legg and Marcus Hutter proposed it as the ability to achieve goals in a wide range of environments. Here we look at the concept of intelligence from the lens of search. And it gives us a unified definition that proves the validity of all other definitions, while encompassing them all at the same time. If all problems are search, solving them means we reach the goals of search. Here, intelligence is the ability to search faster, better, and more importantly safer with regard to for how much resource has one spent.

That is exactly why we think! The idea is, instead of spending the actual resource to search. We could practice the search in a simulated environment that consumes less resources before we actually try it out. That is the reason why the brain is the most useful organ. It allows us to think, to understand and predict the results before we actually act, which could save a lot of time and energy. Because thinking is economical, that is why we associate thinking as a trait of intelligence. But thinking is not the only form of intelligence. We could let someone or something else to think for us.

Given the limitations of the brainFoster, J. K. (2009). Memory: a very short introduction (Vol. 194). Oxford University Press., such as its faulty, everything we remember can be forgotten, or even, for the worst, altered, it would be nice to have something that relieves the brain from doing the search all the time. It is the same reason why we don't ride horses anymore. Because cars are much more reliable! It's the very reason why we invented computers…

∂ Can we create something that thinks?

Here you might think that with the concept of the Turing machine, all the past scientists did was perhaps implementing it? Wrong! From the Turing machine to a computer was not a simple path. In reality, it's reversed. Turing was born after the first general purpose computer had been designed.

You might have heard about abacuses, calculators, differential machines. They even existed long before Turing did, but they did not have the capability of thoughts, they were not Turing machines. All they could do was some basic arithmetic functions. They could, in theory, simulate some thinking, but not all of it. Anything that is equivalent to Turing's hypothetical machine can do much more, because it's the blueprint of mind itself. Humanity had to wait until the early nineteen century to see one.

It was Charles Babbage, the father of computers, and Ada Lovelace, the world's first programmer, that presented the first Turing-would-approve machine to the world: Analytical Engine. Programmable using punch cards, it was the first machine that was a Turing machine even before Turing himself understood the concept thoroughly, even before he was born! It's not surprising that hardware usually came before theories, the story of computers was also one of such things. Seem like people did not always realized what they had unleashed upon the world.

We now understand that computers are simulators of thinking, and from the previous sections, thinking is a simulation of search. And only those computers with Turing's approval could do any kind of search that we could. Nowadays computers have become much more than just tools for search. They are simulators, for any purpose, playing music, games, meeting people online, etc. Though, these things do not seem to have anything to do with search, it's always in people intent to search for something that entertains them, and computers as simulators not just happen to serve those purposes, they were meant to help you find them!

Now that we have hardware and theory, if computers are built to help searching for answers, but how? We could use computers to randomly try actions and generate results for us, though faster, but that is not efficient, not intelligent at all. What follows is that we can improve the search, if we look at each problem individually? Each problem has different search nature. You may know how to search for a toilet in a department store, but you can't use the same search strategy to win a lottery, or design a best device for the disabled. One can't always employ the same strategy and expect to succeed anywhere. And this is what computer science is for.

The catch here is while other disciplines teach you how to do search to solve problems in their fields: physics, mechanical engineering, economics. Computer science studies how to digitize those problems into computers, and how to use computers to search for the answers efficiently, regardless of the fields or disciplines from which they originate.

If you are reading on a computer, be aware that the thing in front of you is a product of generation of great minds in the marvelous endeavour to understanding ourselves. It is totally fine to feel both awed and excited as much as I felt when I wrote this article on a laptop.