According to Tegmark Max, professor at the Massachusetts Institute of Technology, intelligence represents the capacity to execute complex and diverse tasks. To him, human intelligence has a very large spectrum while artificial intelligence spectrum is narrower. But, lately, artificial intelligence has become more and more developed in our society. In the 1960’s, the first program able to solve Turing’s test was created. This test consists in trying out machine’s abilities to show a kind of intelligent behavior, equivalent or not to a human. If we cannot distinguish that we are talking to a machine, the test passes. That was only the beginning. In the past years, artificial intelligence is going further: it can make a better diagnosis than doctors, it won the game of go in 2016 against Lee Sedol, the world’s number one player. Today, artificial intelligence is developed to the point of being able to interact with people. The growing development of robots, which are more and more intelligent, is a blatant example of artificial intelligence development. Taking as example those robots in charge of helping children affected by autism, what does future holds for human intelligence, considering AI development? More specifically in the medical area, what position man and robots may occupy in a few years?
While artificial intelligence is expanding, and the tasks that it can solve are increasing, we may wonder what still remains to human. To Larry Tesler, the computer scientist who invented the function « copy – past », human intelligence is and will remain « what machine doesn’t accomplish yet ». This idea was also developed by the robotician Hans Moravec who saw that the tasks which are difficult for us are the ones which are easy to perform for a computer, and vice versa. This leads to the Moravec’s paradox : when the human brain has perfect control of a task, it runs unconsciously, unlike the poorly controlled ones, and they seem easy to us. The example used is the following: it’s easy to make a computer as strong as an adult to intelligence tests such as the game of go or chess but it is really difficult or even currently impossible to imitate children behavior when it’s about perception and mobility. Knowing that, how can we imagine letting our children watched by robots? We have to know that autistic children have their own perception of the environment. So, if a robot cannot assimilate children behavior, especially autistic ones, how can they assist children? At the current stage of development, they can only be considered as a support for human and medical teams. In the case of children with autism, medical teams rely on new tools like: tablets with adapted applications, robots, virtual reality; each can be used in specific contexts.
Moreover, many experts such as Laurence Devillers, researcher at the National Center for Scientific Research, who is also a member of the French Research Commission on Ethics of Research in Digital Science and Technology, thinks that IA has no comparison with human intelligence. To them, we have to stop thinking that « super artificial intelligence » we are building today is able to take control over humans. According to Laurence Devillers, we have to demystify robot’s capacities, « because man has a natural tendency to anthropomorphize the machine and give it capabilities that it does not have. The robot is much stronger than us to do calculations and has skills that we will never have, for example to hear ultrasound and infrasound, or to hear better than us if we are old, but robots will never know the taste of an apple, will never want to pick an apple let alone eat one! Except by imitation! ».
Indeed, artificial learning is connected with imitation and association. Today, robots deal with imitation. If we want them to substitute a human, or even to assist a person, we must teach them how. If we go back to children behavior, we know that their reactions are unpredictable: in the same situation, two children will have two different reactions. It’s even truer for autistic children, with which even human or experts might have trouble into predicting their behavior. So, how can a robot do it? Even if, one day, robots will take one important place into autistics lives, by helping them to interact. Yet, they cannot replace the medical staff that are daily beside those children. Moreover, if there is too much artificial interaction around autistic children, how can we be sure that they will not develop a kind of addiction to artificial tools, and might also push them to become introverted.
According to Norwegian studies, the intellectual quotient decreased among generation born after 1975, and especially in Northern Europe. Calculated by comparing the real age of the child with his mental age, this method aims to assess individuals’ abilities to deal with certain problems such as reasoning, thinking, or understanding complex ideas. To explain this decrease, some arguments are highlighted: Richard Lynn base its arguments on genetics reasons with the dysgenic effect, while some studies show the impact of endocrine disruptors and the cultural environment. But let’s focus on the last point. The omnipresence of screens in our everyday life leads to visible brain changes, with the decrease of the gray matter in certain areas of the cortex, particularly these links with interpersonal communication and understanding. Today, books become shorter, magazine and newspaper articles become simpler, cartoon pictures and television become more prevalent, and entertainment replaces reflection and debate.
On the contrary, reading leads to multiple neural networks and increase intelligence, and it is probably the learning of reading during the 20th century that has increased IQ. But today, as we can see in the graphic, young people are reading fewer and fewer books and this trend tends to continue. Aforesaid, even medical staff may not be able to understand autistic children. What the situation will be if, with the decline of IQ, things become worse and worse? If that’s the case, are robots intelligent enough to replace human? Nowadays, it isn’t possible, but what if artificial intelligence still continues to increase?
Going back where all started, in 1956, « artificial intelligence » struggled to develop and win, until 2000, with the burst of the internet bubble. The AI is supported by a new wave, which lasts until today. In twenty years, progress is significant: during this period, we went from simple algorithms to robots that can replace humans in some of these tasks. If we encourage this moment, can we be convinced that man will not be replaced by machines … According to Arend Hintze, professor of integrative biology and computer engineering at the University of Michigan, we can distinguish four types of artificial intelligence: the reactivity, the limited memory, the theory of mind and the self-conscious. Today, only the first type was possible to realize, and they are still working on the second one. But what will it be like when we are in the final stage of this process? Which one will be man’s position? Already today, robot artificial Intelligence are able to be reactive and to interact. Leka robot, for example, is a small spherical robot designed to help children with autism and other neuro – developmental disorders to learn by having fun. Some studies point out that using wisely, robots are better than human to assist autistic children. Robots help them to become widespread.
Garry Kasparov, the chess champion who was beaten by IBM’s Deep Blue supercomputer in 1997, addresses in his book Deep Thinking, the two sides of AI. A few years before his defeat against the AI, he said « I don’t know how we can exist, knowing there is something mentally stronger than us ». But after turning every detail of his defeat, he turns out his point of view about new technologies: using computers for « the most ungrateful aspects » will free us and it will allow us to devote our intellectual faculties to « creativity, curiosity, beauty and joy ». If we don’t want to be machine too, we have to seize this opportunity.
Today, most people see the rapid growth of IA as a threat, such as for example with job losses. However, we have to think in terms of synergy, and not in terms of competition. Paul Daugherty and H. Jales Wilson, Accenture Leaders, wrote in their book « Human + Machine », that AI will increase human potential. In more and more companies, humans and smart machines collaborates ever more closely: companies can change process on the fly or they can completely reimagine them. Today, AI is changing the rules of how companies operates. Still according to them, artistic people will have a huge place in their model: they will have to « design, adapt and manage » services robots to the customers. At their side, some professions as comedians, empathy and human conversation are needed to be able to meet client’s expectations. And this is exactly what robots does: they allow to reduce medical teams on charge of easy tasks and to free up time for tasks with higher added value which aren’t actually easy for a robot. Human intervention will appear in a second time to analyze child results and evolution. Moreover, autistic children have specifics needs such as the need to interact with someone. Therefore, robots and human may capture children’s attention. In order to do this, the robot design is essential (such as artistic people).
Diseases removal, suffering, and disability, increase our physical and intellectual performance, overcome old age and death… All these points are the aspirations of transhumanism. According to Max Moore (1990), « Transhumanism is a class of philosophies of life that seek the continuation and acceleration of the evolution of intelligent life beyond its currently human form and human limitations by means of science and technology, guided by life-promoting principles and values ». As an example, Kevin Warwick, a British scientist and professor of cybernetics at the University of Coventry in the United Kingdom, got grafted, in 2002, electrodes in his nervous system to interact with computer and machine. Considered as the first cyborg of the history, Kevin Warwick could then take control of a robotic hand, with the help of any Wi-Fi network (used to be connected with the hand). A few years later, he grafted electrodes in his wife arms. This allowed them to interact in the form of telepathy, connected by nervous system. As we know, interaction with autistic children is difficult: maybe this type of interaction is a new solution. We use technology and artificial intelligence, without losing the human side of interaction.
Aforesaid, artificial intelligence improves tasks solving. In short time, we might pass from the creation of internet to machine learning without anyone knowing the limits of artificial intelligence. IA must not be seen as a threat, but as an opportunity to help people. In the case of autistic children, robot cannot replace human: they aren’t developed enough and cannot integrate their specific behavior. Human play a key role. Also medical teams are using robots and artificial intelligence to help themselves in some sectors: communication, social interaction. This method gain ground and shows that the collaboration between human and machine increases autistic children well-being and behavior.
As if I have chosen to work on this example, robots and artificial intelligence are tomorrow serious challenges, and this, in different business sector. To conclude, as Max Moore said: « if you want to change the word, you will not do it as a businessman, or as a politician, or as an athlete. You can change the word by being a scientist or an engineer ».
By Clarisse Bouet, student in M2IESC 2019-2020