Tech experts are rapidly changing our society by the development of systems driven by Artificial Intelligence (AI). Our brains try to adapt to this evolving world, although not without consequences. How is our behaviour affected by the urge to adapt to new technologies? And should decisions for technologies with an impact of this magnitude be taken solely by a couple of commercially driven tech experts?
Once I was a scuba diver in the sea of words. Now I zip along the surface like a guy on a Jet Ski”. With these exact words Nicholas Carr (‘one of the blogosphere’s most prominent, and thoughtful, contrarians’) asked if “Google [is] making us stupid” in his 2008 Atlantic article. This influencing article, published by the American magazine: the Atlantic, was coined as a once in 50 years ‘ intellectual grenade [lobbed] into our culture’. The predecessor dates from 1945 and is considered the blueprint for the world wide web.
Back in 2008 Carr noticed his (and his environment’s) attention span decreasing through the use, and through the very design of the internet. An example is the extensive use of hyperlinks, which ‘propel’ you to other sites. Ever since, research has indeed shown how using the internet has influenced our behaviour. It has been proven that it has a negative impact on our attentional capacities, memory processes and social cognition. Also, researchers have found a new phenomenon called Google-amnesia, where people do not even try to remember facts they believe can be found on Google. Other experiments have shown as well that people tend to store information found on the internet only in their temporary short term memory.
This is not to say that this is a bad thing per se. Perhaps we are merely adapting to a new technology. This has happened throughout our whole history, and every new technology has faced numerous critics. Socrates himself famously opposed the written word, as “words themselves are not a complete representation of knowledge, but rather words are to knowledge as pictures are to their subjects”. While he was undeniably right, Socrates failed to see the complementary potential of the written word, which could ‘spread information, spur fresh ideas, and expand human knowledge (if not wisdom)’. So Maybe Carr is just another Socrates, failing to see the full potential of his use of the Internet.
So Maybe Carr is just another Socrates, failing to see the full potential of his use of the Internet
However, never before in history were technologies adopted so quickly or was the rate of invention so high. Nowadays when an algorithm is created the only thing left is to upload it to the web, where it is immediately available for billions, whatever the ethical considerations for that invention might be. One big catalyst here is Artificial Intelligence, as AI is able to seamlessly use algorithms and data that has already been produced.
With the pace of the development of AI systems, companies have a great responsibility. Not only do systems have to work well, they also must be ethically proof. Nowadays this is still not always the case. Although Big Tech companies do have ethical committees, these are often overruled by monetary decisions, or even seen as a diversion of attention. After all, almost every ‘Big Tech’ company is a capitalist company, which has to report to stakeholders, and is always split between the inside pressure for money and the outside pressure for ethical products.
As shown above, the fact that algorithms have changed our behaviour is without question. How algorithms will continue to change our behaviour is what remains to be seen. It is therefore interesting to look at how AI will impact three unique human skills: sociality, creativity and empathy. This is what the main focus of the rest of this article will be about. It is without doubt that the very design of new technologies can impact the life of billions. Therefore it is also right to ask where the power should lie over those technologies: with the select group of caucasian middle-aged males who produce it, or with the billions who use it?
One characteristic that distinguishes us from many other animals is our sociality. Some researchers even argue that social skills are more critical in developing intelligence than the ability to use tools – which is often presented as a sign of intelligence. Professor Robin Dunmar came up with an influencing hypothesis (The Social Brain Hypothesis) that basically tells us that social species tend to have relatively larger brains than solitary species. Although that doesn’t mean that being social directly makes one more intelligent, there is also evidence that social species (closely related to solitary species) are better than solitary species at cognitively challenging tasks. Therefore it seems sociality is important for our intelligent behaviour. Nevertheless the urge for social skills is vanishing since we are communicating more with autonomous machines designed by a few whizz-kids programmers. What could be the impact of substituting social interaction among humans with human AI interaction?
In an article from the Monash university in Melbourne the author argues that once people don’t use a certain skill, they will de-skill or forget it. The article mentions sociality as one of these skills. This can cause some serious problems. To elaborate, the author explains a real situation where customized teaching systems were used in a school. All kids got a personal AI teacher, which had the advantage that every child could be taught on their own level and at their own speed. This sounds fantastic, but the interaction between the children in class was drastically reduced. A mother reported that her child wanted to bring earplugs to the class in order to fade out the noise of the other kids. And the social consequences of these devices did not stop there. After a while parents started withdrawing their children from these schools because (according to them) the replacement of the human teacher led to depression, anxiety and fatigue. This situation clearly shows how the use of personalized AI systems affects our sociality dramatically.
The problem with these chatbots is that they are designed by commercial imperatives, […] but these applications do not let us self-reflect honestly or show us a painful truth,
However, some might argue that AI gives us opportunities to practise social communication in a safe environment (with a non-judging AI system to talk to). A common example is in communication with chatbots. Currently some chatbots are designed to be both your friend and your therapist. If you are struggling to share your emotions with other people, such a chatbot could be an outcome. But is this a solution for your wellbeing in the real world? The chatbot might give you the opportunity to feel better quickly, but it does not communicate the way humans do with each other. The problem with these chatbots is that they are designed by commercial imperatives, in other words to deliver you short term happiness. But these applications do not let us self-reflect honestly or show us a painful truth, because a user might just delete the application after such a confrontation. Without being criticized by a chatbot we might become more self-centered & less caring about the people around us.
These examples show that AI devices are already having an impact on our social behaviour. This trend may continue since the use of personalized AI systems (such as chatbots) is increasing. A continuance of this trend might cause the near complete disappearance of human-to-human interactions for some in our society. And this is not only happening in our phones or on the internet, also in healthcare the use of chatbots is considered increasingly.
Social behaviour appears to be a causative aspect of intelligence, as we’ve seen with the experiments between social and solitary species. Consequently, we should not think lightly about the development of systems that undermine social human to human interaction. Finally, as the discussion about chatbots showed, the effect of AI systems on social behaviour is not always clear by the design of an application. However, it is clear that applications designed by commercial imperatives do not guarantee ethical proof systems. Currently mostly tech designers are in the loop of designing these systems, while these devices have the potential to change the behaviour of almost everyone on earth.
Another important human characteristic that is currently tried to be simulated by AI systems is creativity. Big Tech companies are investing in algorithms capable of looking at examples of poems, paintings or music pieces to regenerate similar kinds of these art pieces. The algorithms cannot only be used to imitate pieces, but it can also be used to learn an artist’s “style”. An example of this is illustrated in the image on the right where a picture of a few buildings next to a river can be drawn in famous artist’s styles. Through this method completely new pieces can also be generated.
In 2017 a group of researchers came up with such a system to generate images by itself. The system managed to create images which slightly deviated from styles learned by the examples of images. After creating the pieces of art an experiment was conducted in which human subjects had to compare images created by humans with the AI generated ones. In 60% of the cases subjects selected the AI images as more novel and more aesthetically appealing. There are also lots of other examples in which AI is generating sophisticated works such as poems or music .
Evidently, AI is currently at the point that it can imitate creativity in some tasks. Of course creativity is much broader than creating paintings, music or poems, but nonetheless other creative tasks are getting automated as well. The key question now thus arises: “How will creative AI systems influence our own creativity?”. If creative tasks performed by AI are more appreciated than the human performance, is there still a widespread incentive for us to use our own creativity? Some earlier automisations suggest that the latter is not the case. By inventing the calculator, people got worse in performing arithmetics. By having spell-checkers on our laptops and phones people do a worse job in spelling correctly without them. Children growing up with these techniques show worse spelling skills than people that grew up without them. So it seems that once a task is automated, our skills for that task slowly vanish. If this is also happening broadly with creative tasks, people will consequently become less creative.
Not everyone sees creative AI as a threat to our creative abilities. Some argue that AI can work very well together with human creativity and instead human will reach a higher level of creativity. For example, we are able to compute more complex formulas with the introduction of the calculator than before. In 50 years we might think the same about creative tasks.
How creative AI is going to affect our own creativity remains disputable. At this moment genuine AI creativity (containing all facets of human creativity) is far from reached. Nevertheless Big Tech companies are interested in expanding AI in this domain. What if we succeed to create truly creative machines? Then creative traits in humans no longer give a competitive advantage to one another. According to ‘evolutionary psychology’ – a research field inspired by Charles Darwin’s theory of evolution – changes in the human mind can occur relatively rapidly (as opposed to his classical evolution theory that is about millions of years). Therefore, from this point of view, it is interesting to see what is going to happen to our brains if creativity becomes more and more irrelevant. Especially since Big Tech companies are directing this evolutionary process.
Commonly described as the “ability to recognize, understand, and share the thoughts and feelings of another person, animal, or fictional character”, empathy is indeed “the ability that makes us truly human” . It is undeniably one of human’s most unique and uniting feelings, which motivates people to show behaviour unheard of in most other species. Try convincing a fox to give up some of its food for a starving nephew on another continent.
Sadly, with the digitalization, there seems to be a decrease in empathy among humans. Important aspects are the decreasing sensitiveness for emotional cues such as eye contact, tone and body language. And also, quantity is gradually replacing quality, with a massive amount of likes outweighing a few real compliments from your true friends.
Apart from these more distant kinds of interactions, a new field of (online) communication is also emerging and taking its place into our society. This form of communication cuts out other humans totally, and is fully powered by AI. We are of course talking about ‘chatbots’. The commercial potential for these applications are endless, such as AI-driven apps consulting with mental health problems, or even taking care of dementia patients. A critical feature for these human-machine interactions will be, you might have guessed it, empathy. Research has shown that artificial ‘empathy’ plays a key role in human-machine communication. Robots that show empathy are seen as more friendly than robots that are not showing empathy. There is no doubt that empathy will be the next big thing for chatbots, as this has proven to result in more likeable and believable chat-companions. One question we need to ask however, is whether we would even want empathic AI?
Social robots and empathizing with social robots may negatively affect our ability to empathize with other humans
Research has shown that ‘social robots and empathizing with social robots may negatively affect our ability to empathize with other humans‘. This thesis states that social robots will have the same effect on our decrease in empathy as ‘alternative facts’ (e.g. the number of people present at Trump’s inauguration), since “social robots remove the need to empathize with other human beings in the same way alternative facts do: by making it possible to have exactly what we want, without compromise”. Of course, not everyone agrees with this view. Global Head of Strategy and Offerings at IBM, Jesus Mantas, claims that ‘empathy, our species survival trick, is the clue to our next evolution’. He states that by enhancing our emotional mind with AI (we have mostly enhanced our rational mind up to now, e.g. calculators), we will be able to enter a society as never before. What if, he asks, our ability to understand is enhanced a hundred-fold? Mantas believes we could finally ‘reverse-engineer’ our unconscious biases, and by doing so finally make a truly ‘common sense’.
While this may be true, team ‘natural-empathy’ has an objection to this point. By increasingly relying on technology, we gradually lose the ability to emphasize ourselves. Empathy is also described as “a muscle: the more you use it, the stronger it gets”. So actively practising empathy is beneficial in itself, as one also trains its own empathy. Another counter-argument can of course easily be raised here: while the above holds, it would not really matter, since we would have the technology available to enhance our empathy anyway. This is of course true, but there is world-wide growing concern for the hackability of these hypothetical future AI-tools. As Black Mirror showed brilliantly in its third series final episode “Hated in the Nation” , the purpose of AI need not be malevolent, it could always be hacked and turned into the unthinkable. What if this were to happen to our most trusted ‘empathic friends’?
Children who grow up relating to AI in lieu of people might not acquire “the equipment for empathic connection”
Additionally, there is another reason why we would not want to leave our empathy in the hands of AI. In 2010 the illness EDD (Empathy Deficit Disorder) was coined by business psychologist, psychotherapist, and P.H.D researcher into adult development Douglas LaBier. He states that if you are ‘unable to step outside yourself and tune in to what other people experience, especially those who feel, think, and believe differently from yourself’ you suffer from EDD. This ‘has profound consequences for the mental health of both individuals and society’. As we rely on AI for empathy we will thus start to behave like something that is noted as an illness today. Also, Sherry Turkle, the MIT expert claimed that : children who grow up relating to AI in lieu of people might not acquire “the equipment for empathic connection”.
As discussed above, the digitalization has already had a significant impact on our empathy, where we are less prone to emotional cues and gradually have started to value quantity over quality. The use of empathic AI in chatbots however will lead to a whole new approach of empathy, where we will over time become accustomed to only talk to our ‘perfect’ empathic chatbots, and gradually lose the ability to empathize without the aid of technology.
As social AI systems can undermine human-to-human interaction, human creativity might be altered by creative AI systems, and empathetic AI chatbots can cause a decrease in human empathy. There seems to be evidence that AI can change our behaviour in some of those features we deem ‘inherently human’.
The common divider for most of these AI applications is that they are still being created by a very selective group of people, mostly within the ‘Big Tech’ circle, while it is evident that the design of these algorithms have the potential to change the behaviour of whole societies. Therefore, the impact Big Tech has on the world is tremendously disproportionate to the few people that are creating it. While we know there is no single solution, it is important for everyone to realise how their behaviour is altered by a select group of people, not representative for the world population.