Let it Learn Machine Learning: Artificial Intelligence Needs Both Logic and Ethics

The power with which the accelerated and unlimited development of artificial intelligence (AI) has further endowed new technology raises concerns that the machine will soon be more important than man, which is why its application needs to be humanized. The computer revolution thirty years ago placed technology and STEM sciences at the center, but now experts from that field are calling for help from humanists, primarily ethicists, as well as those from social sciences, because ethical discussions about artificial intelligence are becoming increasingly relevant.

Humanities in general, and not just philosophy, directly address the question of ethics and the ultimate goal of human life. Since artificial intelligence has entered almost all areas of life and is becoming capable of replacing an increasing number of professions and leaving people unemployed, but also monitoring and manipulating data, there is a danger of undermining democracy, which could have catastrophic consequences for humanity in the long run.

The renewed search for a broad and fundamental understanding of ethics in the domain of artificial intelligence, with radical implications for the restructuring of social power, was recently described in the book ‘Atlas of AI’ by Kate Crawford, a professor of communication and technology at the University of South Carolina in the United States and a principal researcher at Microsoft Research. Crawford writes: ‘We need to focus less on ethics and more on power because AI is always designed to amplify and reproduce forms of power.’ She argues that we need to examine the structures of power into which artificial intelligence is embedded in terms of ideas of equality, justice, and democracy, fundamental ethical values of democracy.

Restructuring Social Power

The main question is how ethical it is to use AI to enhance human performance and avoid failure, and where is the boundary between man and machine? The decisions that will be made will determine the future of civilization, and they need to involve not only the humanities but also the social sciences.

—

—

For example, lawyers will need to determine who will be responsible if the AI engine that powers a self-driving car makes a mistake; will it be the company that designed the engine, the one who taught the car to drive, or the artificial intelligence? Ethics should help prevent straying from this.

Recognizing Abuses

The importance of using the humanities, especially ethics, in the development of artificial intelligence was explained by Prof. Dr. Senka Maćešić, a full professor at the Faculty of Engineering in Rijeka, and vice-rector for digitalization and development at the University of Rijeka.

– I believe that, just like with all human inventions and discoveries, we must first focus on preventing abuses and harmful uses of artificial intelligence algorithms. For example, if we want to use AI to select candidates for a job application, and we use data from previous employments that contain biases against certain social groups, AI will reproduce those biases in the employments for which it will be used. We know that AI is used to manipulate political views and election results in otherwise democratically organized countries. We know that it is also used for ‘classifying’ consumers, in order to create dependency on social networks and achieve huge profits from advertisements. All of these are already present abuses that need to be recognized, legally regulated, and prevented. However, if AI is used to predict glacier melting or accurately diagnose a disease, it is very likely that such AI will bring only benefits – explains Maćešić.

Preserving Humanity

The far-reaching consequence of uncontrolled application of artificial intelligence, which, according to experts’ predictions, will mark the next thirty years as the internet marked the last thirty years of development, will be unlimited manipulation of society if the views of humanists and experts from other fields unrelated to STEM and computer science are not taken into account.

– Various abuses of artificial intelligence could primarily lead to societies where control and manipulation over individuals are unlimited, reaching completely inhumane and absurd levels. It could be societies where, under the guise of care, optimal guidance of personal potentials, and assistance in various choices, people are actually managed in subtle but also extreme ways. A future could be imagined that would lead to changes in independence and even decline in some basic life skills if we overly delegate them to personal assistants based on artificial intelligence algorithms – emphasizes Maćešić, who believes that we will avoid such a future.

She believes that there will be something that is a more instinctive tendency of most people to remain what we are, to preserve our humanity, even if it seems outdated. However, this tendency can only be realized with the help of the humanities, clearly expressed, and can thoughtfully and systematically help shape our future.

Questions for Answer

Professor at the Faculty of Humanities in Rijeka, Prof. Dr. Elvio Baccarini, who deals with the philosophy of ethics, argues that although AI, like other advanced technologies, has great potential to influence the shape and quality of human lives, and the possible contributions to progress are significant, it is essential to consider the implementation of technologies in personal lives and in society in general in accordance with reasonable value, moral, political, social, and anthropological judgments.

According to Baccarini, there are several questions that need to be answered: How does the implementation of AI and its increasingly pervasive presence in our society and lives affect individuals’ cognitive abilities? Does it modify and in which direction emotional and affective potentials? How does it affect the abilities of empathy, affiliation, and identification with the common good? Does it shape our political orientations and influence our ability to make autonomous decisions? Does the implementation of AI affect the fairness of social arrangements, promote greater social equality, or exacerbate the social divide? Which applications of AI favor human virtues, quality of life, social progress, and strengthening of justice systems, and which threaten them?

– These are some of the questions that determine the direction of AI development and implementation. The humanities and social sciences are called to address these very topics. Ignoring the humanities in the advancement of new technologies can lead to development that is only partially directed towards the common good and various values. However, besides engaging the humanities in planning, developing, and implementing AI, it should be kept in mind that these are disciplines that rarely provide final and unambiguously correct answers. For this reason, it is essential to engage the humanities and social sciences as expert support for the democratic process and a system for correcting wrong and rationally poorly set thinking, and directing attention to relevant anthropological and social values – explains Baccarini.

He adds that social and humanities sciences should act as a source that offers guidelines to politically organized society, although they are not always unique and accepted by general consensus in the professional community. Furthermore, they should provide professionally shaped arguments about the value contents at stake in democratic processes, which citizens need to weigh and reasonably adopt, assuming that citizens are conscientiously and reasonably engaged in democratic processes.

Multidisciplinary Approach

Prof. Dr. Jan Šnajder from the Faculty of Electrical Engineering and Computing at the University of Zagreb argues that thinking about the development and use of artificial intelligence should be approached multidisciplinary, meaning that both the humanities and social sciences, as well as all other sciences related to the area in which AI is used, should participate.

– Now the ethical component of artificial intelligence is beginning to be actualized because there was no awareness of its necessity at the beginning of its development. It could happen that technology advances faster than society, that we develop AI that is too good, while as a society we are not ready for it. This is a real danger that is hard to eliminate. Social and humanities sciences are relevant and will help us avoid slipping into technocracy – emphasizes Šnajder.

He argues that a society without humanities and social sciences is doomed to failure and that philosophy of ethics and mind, psychology, sociology, anthropology, neuroscience… should participate in regulating the power of artificial intelligence… because all these disciplines could jointly solve the problem of the human mind and cognition. Depending on the area in which artificial intelligence operates, if it is in medicine, then medical experts should be included in the regulation, and this rule should be applied to all other areas.