Does intelligence imply absolute rationality ?

 If you look up ‘Intelligence’ in the dictionary you are likely to find a definition that goes something like this: “The Ability to acquire and apply knowledge and skills”. This seems fair. Certainly being able to acquire knowledge, and apply skills, are without doubt requirements for an entity to be spoken of as intelligent. There is nevertheless a problem with this definition of intelligence, and it is simply this: It’s telling you that intelligence is an ability, but it isn’t revealing anything about its true nature. The skills and knowledge that we are able to acquire are simply tools, for coming up with the best possible decisions. Therefore, I would argue that this definition clearly implies a link between intelligence and rationality.

Despite the title, this article won’t reveal you the true nature of Intelligence, unfortunately that is yet to be figured out. What you will find in this article is an attempt to explore the significance and limitations of rationality in the context of intelligence.

As a society we draw a strong link between intelligence and rationality. The ability to acquire knowledge and skills is an extended tool for considering all possible variables and then making the most logical and rational decision. If we think of intelligent people, we think of great physicists and mathematicians, Einstein always keeps sticking out his tongue at us. Without a doubt, the ability to think and act as rationally as possible, is an invaluable skill for a human and a cornerstone of any theory of intelligence. However, what if we expand this definition to a theoretically absolute form of rational entity? The philosopher Nick Bostrom has published several articles on Artificial Super Intelligence (ASI). He argues that technically, this type of supercomputer could outshine humans in any field. It could follow the most rational and logical approach to any task and would be able to acquire additional knowledge and skills in order to keep on pursuing its objective. So, according to the Oxford dictionary, an ASI would completely fulfil the criteria for the highest form of intelligence. However, Bostrom raises a crucial point: To exercise its intelligence, the ASI would always need a clearly defined task, in relation to which it would make its choices. Bostrom mentions the example of an ASI that has been tasked with producing paperclips. With its superior rationality and skills, the ASI would keep producing paperclips until the whole world is awashed with them, in the process of which it would destroy all of humanity and eventually itself. This begs the question, is this still a superior form of intelligence, or does Einstein still stick out his tongue, even at the most sophisticated machine?

So if we cannot speak of intelligence in terms of the most rational behaviour then perhaps we could speak of it in terms of irrationality. Irrationality is amongst the many sets of skills that sets us apart the from machines. Irrationality is a component of human emotion, something that makes us feel alive and defines us all together. An excellent example of irrational/emotional intelligence can be passion. Passion is an irrational beacon of light that guides us through a significant amount of aspects of our life. Passion can be purely irrational, like one for a sport, or they can be a driver that motivates a student to pursue physics in order to fulfil his desire to understand the Universe. Now let’s consider him. He was not the best in his math class in high school, and he didn’t excel in most of the other subjects, he was however the one who was never satisfied with his grades and would always stay late after school to improve them. All because of his dream was to study the universe. While we as emotional beings can connect and empathise with this behaviour, yet it is purely irrational. However, taking a rational approach in these situation would recommend him to not focus on maths, and focus on easier and less time consuming subjects, which would result in him pursuing a less challenging subject in university, and ultimately, a less challenging path in life. But would we say this is intelligent? I’m sure that almost all of you would say no. In the world we have many examples of people who pursued their passions, making every irrational choice possible and managed to accomplish their task, and surely we consider these people as intelligent, sometimes visionaries. Look at the story of successful entrepreneurs. In the case of Amazon founder Jeff Bezos, he had studied in finance his whole life and already had an accomplished career, yet he decided to give it all up. He then endeavoured to accomplish his vision for his start-up, and just over twenty years later, it became one of the largest companies in the planet. The result of this is that today, by most people’s standards, Jeff Bezos is considered an intelligent person. Does this mean that the true meaning of intelligent is to follow your dreams and always act irrationally? Unfortunately, no... and I’m happy to tell you why. An irrational approach to intelligence also implies a different attribution of causality. This is because, while in the rational approach people are considered to be intelligent in terms of the means with which they start (their ability to acquire knowledge and ability to apply skills), in this irrational/emotional version of intelligence we are speaking of intelligence in terms of the end result. We say that AI is intelligent because of its ability to process immense calculations in a few seconds. But we do not judge Jeff Bezos to be intelligent because of his ability to make start-ups that will then become multi-billion dollar companies, but we think that he is intelligent because he managed it with Amazon. Just as we previously inferred that pure rational model of intelligence is incomplete, we can state that the irrational model of intelligence is incomplete because we judge these examples as signs of intelligence retrospectively, so we only consider them as cases of intelligence if they are successful. To put it more simply, you might scream ‘you’re a genius!’ to your friend after he won 20 quid on a scratchy, but you would tell that to him if he’d lost.

What connects Bezos, Einstein, and your mate with the scratchy, what distinguishes them from the supercomputer is that they do not have an absolute purpose. While they might be devoted to one specific task, and heavily use their rational intelligence to obtain it, the same intelligence would prevent them from flooding the world with paperclips. This implies, that despite the computer’s absolute rationality, there must be more elements to human-level intelligence. The fact that often seemingly irrational behaviour can be perceived as intelligent, suggests that intelligence can not be viewed under the terms of rationality alone.

This article was written by David Neppel and Fillipo Ricciardi