JYes, I had decided to delve a bit deeper into human and artificial intelligence, and as I was reflecting on this, I remembered a great thinker from the 13th century. In the time of Thomas Aquinas, a figure emerged who experienced enlightenment, and this is fascinatingânot only for the content of that enlightenment but also for what he thought he could achieve with it. That figure was Ramon Llull, born on Mallorca. He was the first to write in Spanish, not Latin, which is quite significant for Spaniards, and the University of Barcelona is named after him. Ramon Llull, a 13th-century contemporary of Thomas Aquinas, and although they never met, itâs often said that the very fact they didnât meet allowed Llull to follow an independent path.
It is truly fascinating to consider what he discovered, or what was gifted to him in this enlightenment. During this experience, he developed a highly complex view on the potential of human knowledge, from which he gained the absolute conviction that it is possible to answer all human questions using just 10 concepts. This is an incredible assertion, which he elaborated in a dense book â which I didnât read - called âArs Magnaâ, the Great Art. However, it turned out that hardly anyone could read or understand this book, so he eventually summarized it in âArs Brevisâ, which I have read. Here, he groups these 10 concepts around a central concept, God, with nine surrounding moral concepts, all arranged in a circle. He added further concepts in concentric circles, creating something akin to a roulette wheel with moving circles that, in various combinations, could provide answers to oneâs questions. Now, I canât entirely fathom this system, but I have gratefully used the first part of his enlightenmentâthe nine concepts around the central oneâto set moral thinking in motion. This ultimately became my own small book, âThe Art of Thinkingâ. And so, with these conceptsânot as a roulette wheelâbut as a way of moving thought, I bring thinking into motion.
Why am I saying this? I believe that with Ramon Llull, we can find humanityâs deep-seated desire to find a way to understand everything. He was convinced this was possible, and he developed a method to achieve this understanding. I find this incredibly exciting. You wonder how his ideas were received later in history. Iâll come back to that in a moment. But my point here is that you could say this man created a kind of early computerâa tool of a very different nature and of a high moral standard, as it was intended to reveal not only absolute truth but a comprehensive one, leading to all-encompassing knowledge. âNothing more is needed,â he says, ânine concepts around the tenth, and that is all you need to reach universal certain truth.â
Of course, over time, this was ridiculed or dismissed, which is understandable. But certain later thinkers greatly admired him. One such figure, a largely self-taught mathematician, was Leibniz. He was genuinely enthusiastic about Llullâs methodâand I believe also about the underlying principle and the conviction that human thinkers could reach comprehensive truth in a relatively simple way. Leibniz, who developed the binary number system, envisioned a system that wouldnât use the 10-digit system we know (0 1 2 3 4 5 6 7 8 9), but rather just two symbols: one and zero.
I donât remember learning this in school. Maybe I forgot it because I didnât like it, but I donât recall. Later, when our children were in high school, they learned how to convert binary numbers to regular ones and vice versa. I then became familiar with it. Today, we all know that this binary system made it possible to create a deviceâa machineâthat doesnât have to use the complex 10 digits but rather operates with just one and zero, yes and no. These values can represent numbers and can be used for calculations. Moreover, you can imagine how it could allow certain questions to be answered by asking whether something is a one or a zero, yes or no, and proceed from there. I find it essential to understand these moments in the history of human thought, as they mark the beginning of the kind of thinking that eventually leads to computers and artificial intelligence.
I read, for instance, that around 1900, David Hilbert, a German scientist, perhaps a physicist or mathematician, expressed a wish at a conference to find a formula that could solve all mathematical puzzles. This reflects a similar wish, a deep-seated desire that, amid the complexity of creation, there exists a fundamental law or formula that contains everything. If you think religiously, you might say, âOf course, thatâs Godâthe all-encompassing âformulaâ from which everything else flows.â But even without this belief, the sense remains that such a formula exists, fueling the desire to discover it. And in Hilbert, we can see another precursor to artificial intelligence.
It is known, of course, that artificial intelligence works with, I call them formulas, they are called algorithms. So when you delve into primal computer science - which is what I tried to do, because I believe that you can only experience what you are dealing with if you go back to the beginning - so when you delve into that primal computer science, you find described the algorithm of Euclid, the famous Greek mathematician who brought us Euclidean geometry. He found that when you want to find from two numbers - which can be small or large - when you want to find from those two numbers the greatest number by which division can be done on both numbers, if you want to find the greatest common divisor then you have to follow a certain procedure. He described that procedure. You can do that, you can know that if you want to find that then you have to divide the largest number by the smallest. If that comes out, if there is a remainder zero, then you immediately have the greatest common divisor, which is then the smallest number. But usually that does not come out and then there remains a remainder and because you assume that there is a greatest common divisor, you know that that remainder, that you could divide the smallest number by that remainder. So if that comes out then with that you would have found the greatest common divisor in a second step. Well usually that's not true either and then you divide that and then you get a result, but you get another remainder and so you go on. So again you divide the smallest number by the remainder and so on until finally the division succeeds and the remainder is zero. Then the previous number that you divided with, the previous remainder is then the greatest common divisor. Well you can think that through, we did not do that in school, we were given that - there is also a formula for it - we were given that and yes, you then simply assume that. What you then do not realize at all - and that is what the philosophy of science should take care of - what you do not realize is that you then actually are a believer, that you then let something happen that you could see through and say: Okay if I just know how to do it then I'm satisfied and I don't need to know exactly why is the remainder so important? When you start thinking that, you take that remainder and divide that little number by that remainder, when you start thinking that, then you may have a little difficulty for a while, to grasp that and eventually you may find out but you may not. Then that's very powerless how you feel then. But actually, of course, you know: if this is the way to find the greatest common divisor, then I should be able to understand it. See and that's human intelligence that you can understand that in certain - and this is most obvious in mathematics - that there are certain areas in the mind where you as a human being can really grasp, really understand why a certain procedure is used. You can use it without getting it, but you should actually have a sense, yes I'm just assuming something because it works, but I could see through it. And that's human intelligence. The more complicated the process, the more difficult it becomes to see through it, and that is the pitfall, there is no escape, there is no escape, there is the pitfall that at a certain point you have to deal with procedures that have so many steps that you cannot expect to be able to fathom every step. You know that they can be put, that they have to be put in order to achieve a certain result, you do it, but that so actually against human intelligence. And I dare to say the following: that we as intelligent human beings are on the way that seeing through certain coherences with the help of human intelligence that we just use them, we are losing that ability, or at least are in danger of losing it, because the computer does everything for us and we have no idea what it is doing. All those thousands of millions of steps hidden in an algorithm that is of course also impossible to grasp, to see through there step by step what is happening. Euclid's algorithm can be understood. You can also grasp that you could build that into a machine which could then perform the various steps exactly - you may not say know - but out of habit, which has been introduced into it. That is still very clear with that algorithm. When you have a complicated problem or you just search something on Google, with that you have called up a certain algorithm, called up a certain procedure, which is so incredibly complicated, that as a human being you can only shrug your shoulders and enjoy the result. Meanwhile, the fact is that with our human see-through intelligence, we shrug our shoulders at procedures that we use lustily - me included - and can't see through them at all. Well of course you could get desperate about that, but as far as I'm concerned, the consolation is that when you see through the basic principle, when you see through how it works, then you have actually already won the victory. And I'm trying to go down a very small piece on the path of that victory with this video. And so I summarize again: There are certain procedures that human intelligence can do where it is possible to understand exactly, step by step, what you are doing. When you let go of it, you just think, it will be all right because it works, and you just leave it for what it is and you don't try to grasp why those steps are the way they are - so what we learned in school in proving theorems in mathematics, there we did that, there a theorem was given and that had to be proved. And if you prove it, then you can still memorize the different steps, But you know of course, you can also understand it. And once you understand it, you don't forget it either. That is human intelligence. When that very basic ability of seeing through certain procedures, when we let go of that because it happens anyway, because it all works anyway, then we are on the road to losing that ability, just as you lose your muscle strength when you no longer use them, those muscles. That's what everybody wants to believe, that you have to go to the gym or you have to go for a walk or you have to go for a run or you have to go for a bike ride or I don't know what in order to keep your healthy muscles, you don't keep them if you sit at home on a chair - everybody believes that. But that it is so that you are also going to lose human intelligence as humanity, when you let go of understanding the procedures - yes really not everybody believes that, but that is my thesis now! We will continue next time!
Human intelligence and Artificial intelligence: The art of thinking by Mieke Mosmuller