Computer says no…

Although the news of the period is probably related to the appointment of Eugen Tomac to try to form a government, I fear that we will discuss the same subject next week and, unfortunately, in the following weeks as well. I take advantage of the approaching end of the school and university year, of various exams, and propose a brief discussion about, as they say, the use and abuse of AI (artificial intelligence). The story has been extensively debated in the last 2-3 years, as software and applications in this field have evolved rapidly. Many industries have adopted these innovations and presented them simply as an intensification of digitalization (which no one contests, although there could be some discussions about how it is done).

This is not the first time that human society faces the huge impact of technological innovations that, once adopted, overturn hierarchies, ways of life, trades and professions, norms of all kinds. Industrial and agricultural machines, trains, ships, airplanes, cars, printing, radio, television, satellite television, the internet, mobile phones, online shopping, smartphones, social networks: whether small or large, all technologies adopted in daily life have profoundly changed social life, from micro to macro. Of course, it is worth noting that the evolution of computers and their applications has brought innovation after innovation in the last 40 years and there has been a real crowding around them of small and large impacts on our daily lives. We have absorbed them all, from the state and corporations to the last household that has accessed the internet and mobile phone networks.

It is somewhat comical that in the past, visionary writers and researchers in the field of future studies assumed that there was some kind of pyramid of development and technologies. As we evolve, technologies become more accessible and cheaper, their use becomes widespread, and the same happens with subsequent innovations, etc. Nonsense… the evolution of the internet and mobile phones shows that you can have, as they say, an outdoor toilet and high-speed internet at home. This must mean something regarding how prepared we are for new communication technologies.

Another assumption of those who some time ago began to study and imagine the future, but still along a slightly Marxist line of thought that technological progress equals a better and easier life, was that robotization and digitalization would eliminate difficult, unwanted, dangerous, possibly unskilled tasks, or extremely repetitive, time-consuming, and boring processes (including bureaucratic ones). Efficiency was supposed to mean a better life. Would the jobs of the unfortunate who lived from these chores disappear? No, because they either had to transform into supervisors of the machines involved in those tasks, or they would focus on the qualitative aspects of a case, once the raw work and repetitive processing of data or documents were to be done by computers and robots, under the watchful eye of the official/operator.

The harsh reality of the emergence of AI is, however, that the first roles (not to say jobs) put at risk are the creative ones, then the human repetitive and bureaucratic ones, while there is still something to be done for people in manual jobs. Practically, AI will replace programmers and writers (thus also the so-called real and humanistic branch of creation) faster and more easily than plumbers and construction workers. Or it will transform a multitude of intellectual occupations, relationships, and work interactions, perhaps also ways of contracting and labor norms.

And because today is Inventor's Day, let’s also touch on the education corridor. Education and scientific research have reacted quite late to these challenges. Practically, students have quickly discovered ChatGPT and other software that have emerged in the meantime. The education system is still debating how to incorporate AI into education and research. The problem is not whether or not teachers are familiar with AI tools. Many are, and anyway, that can be resolved. The problem lies at the level of principles and regulations. Let’s remember how many years it took to regulate plagiarism. The major difference between plagiarism and AI products is that, in the first case, if you could show someone black on white, on paper or online, the material from which they plagiarized, by identifying AI writing you cannot present an objective proof that AI was used. You can only present a report made by software that detects AI products and which claims that the respective material was written by AI. But you do not have an actual proof, such as identifying the source from which it was copied in the case of plagiarism.

So the first step would be a guide of good (and bad) practices in using AI and, even if we specify good practices, to foresee what responsibility the person using AI tools has in this process, whether they are using them legitimately or not. But quickly, because otherwise some types of usage will generalize over which we will lose control.

Whether we are talking about education, research, engineering, services provided by private companies or public structures, about administrative procedures and public order, the incorporation of AI into activity can only be done by shifting the focus from the operator's workload to responsibility. If we use digitalization and AI to blame the robot, only then does the real crisis of work in the face of technology come. The increase in legal accountability of organizations and workers who use AI is the true meaning of digitalization – a digitalization that helps and streamlines and which does not become an excuse for organizations that want to escape salaries, nor a Kafkaesque reality for the client who no longer knows who to talk to when they have a problem with the services provided. You know, like that lady from public relations in Little Britain, Carol, who calmly but unempathetically responded to any request from the beneficiary with "the computer says no…".