Artificial intelligence welcomes humans

This focus served as an inspiration, at the end of July 2019, for the round-table discussion “From Socionatures to Humanoid Robots: Transdisciplinary Perspectives on the Relations between Humans and Non-Humans”, organized by the Freie Universität Berlin (FU Berlin) and Mecila (Maria Sibylla Merian International Centre for Advanced Studies in the Humanities and Social Sciences Conviviality-Inequality in Latin America). Held at the Goethe-Institut of São Paulo, the initiative partnered with the German Centre for Research and Innovation São Paulo (DWIH São Paulo).

Among the panel members*, the event was enriched by the international expertise of Prof. Christoph Benzmüller, from FU Berlin, who granted an interview to DWIH São Paulo on artificial intelligence – from its initial facets on behalf of humanity, through scientific innovations with respect to the technologies of humanoid prototypes, with an emphasis on human and non-human relations and experiences, as well as ethical-political implications that bear on the subject. The challenges, dangers, perspectives, technological trends in various areas, rights and duties surrounding this subject were also on the agenda. And one issue stood out, clearly: AI has already invaded our lives and it must be subjected to the spotlights of debates and discussions, asap, so that the transition is beneficial to humanity, since the speed of technological development surpasses human development and limitations. From the perspective of Benzmüller, who currently dedicates his time to researching methods and procedures for regulating the ethical-legal control of AI systems, these technologies come with negative and positive points; but he states that AI holds enormous potential for positively changing the world.

• – In your opinion, what is the current role of artificial intelligence in our lives, especially in human and non-human relations?

  Artificial intelligence is already invading the more remote and intimate areas of our lives. Alexa, Siri & Co. are comfortably ensconced in many living rooms and bedrooms of our homes, listening to everything that happens there. Even in computer games, social networks, or the digital sex industry, human and non-human relations are already merging. This interconnection between the virtual and real worlds, driven by ever improving AI technology, will continue to gain strength and that, of course, will impact our social behavior. We will also experience real partnerships between humans and machines – those trends cannot be seen only in Japan. That development, however, presents both positive and negative points, and the discussion of their implications must begin asap.

  Accompanying and caring for the elderly, for example, present huge challenges for many countries. Robotics, as well as AI, can be highly beneficial, if used with good sense. However, problems could arise through erroneous and highly humanized expectations, with respect to emotional intelligence or the competence of humanoid robots. These technological applications of AI must not be moved and controlled merely by economic interests, but molded by well-founded political decisions.

  – Humanoid robots are increasingly taking their place in social life. How do you view this scenario? What can be expected from this new reality and from the relationships it creates?

  Some nations can hardly wait to grant civil rights to humanoid robots. Saudi Arabia moved forward with this, in 2017. In terms of marketing, the move certainly seems to be very attractive. However, this development, as premature as it is, also has its positive side: such actions drive the debate regarding to what degree, in what way, and with what obligations and rights intelligent machines can and should be integrated with the existing social and community structures. This debate must begin now, since current developments in the field of artificial intelligence are accelerating, while the capacity and speed of social structures reach their limits more quickly.

  In the context of this discussion, other unresolved issues will also appear. In many societies, for example, women still do not have the same rights as men. Therefore, to be concerned about civil rights for humanoid robots seems quite strange. I especially see challenging questions regarding the duties and credibility of future autonomous AI technologies, as well as issues related to the socially acceptable and sustainable integration of such technologies with the existing social structures.

• – How does artificial intelligence affect our routines and our lives? And how can we best deal with this?

  The impact of these AI technologies on the routines of our lives will continue to grow. With the use of AI technologies, our daily activities are not only easier to plan and to manage, but they can also be traced down to the smallest detail. Those combinations will probably leave us increasingly exposed to manipulation attempts, from perfectly personalized advertising to gadgets that, intentionally or unintentionally, consciously or unconsciously, will “help” us in our daily decision making.

  – What are the ethical-political implications surrounding this subject?

  As I see it, the need for developing a trustworthy AI technology in critical application areas is not an option but, rather, an obligation. In this regard, I think the position being taken in Europe “trustworthy AI made in Europe” is great. AI technology has huge potential for changing the world in a positive manner – I myself entered this area, fascinated by this very subject. The positive potential of AI technology must be explored and utilized. However, AI technology also comes with risks, and experience teaches us that from a social perspective, a healthy balance between benefits and risks usually can only be achieved rather unsatisfactorily within its own sphere. The political realm, therefore, must intervene and, in conjunction with industry, develop overall conditions for the creation of AI technology that is socially desirable. It is important to minimize the harmful potential and maximize the benefits for society. Therefore, the development and application of AI systems in some areas should be consciously either discontinued or advanced. I, personally, think, for example, that it is very questionable to develop autonomous lethal weapons systems, totally automated financial markets, state operational or surveillance systems, etc. The easy manipulation of voter opinions, via the social networks, also presents a significant social risk – democratic structures seem to be increasingly under pressure.

  – In your opinion, what are the main areas in which artificial intelligence offers benefits, and why?

  That’s hard to say. The range of applications for AI technologies is enormous and covers many important areas: health, transportation, environment, finances, industrial production, army, corporate management, administration, etc. In each of these areas there are many possible special applications for AI technologies. These technologies will interact with people in many concrete situations, or even totally substitute humans. In some specific areas, AI technologies will be much superior to human beings, and the work processes and structures will be adequately adapted.

  There are various benefits, ranging from increased productivity and efficiency to the reduction of accidents (activities considered to be dangerous can be increasingly transferred to robots). But there are also risks. Mainly the transfer of broad-based decision-making to increasingly autonomous AI systems must be treated with all due caution. Especially in application areas where collateral damage could occur, I would prefer not transforming AI techniques with obvious weaknesses (like a lack of reliability, comprehensibility and transparency, high capacity for manipulation) into concrete applications very quickly. It is usually up to society to clean up the pieces and the damage can even be irreparable.

  For that reason, I am currently researching methods and procedures for regulating the ethical-legal control of AI systems, which I consider to be absolutely necessary in especially critical AI application areas.

  Another challenge arises naturally from the issue of how society’s adequate participation in technological progress and in economic success can be guaranteed. With the growing global concentration of technological power, the answers might not be easily found.

  The exploration of the whole range of applications for AI technologies is far from achieving its peak and both companies and political entities currently lack adequate competence. That is, I deliberately use the plural form “AI technologies”, because it is not merely mechanical learning that has many application areas. The media also seems to have recently backed away from an excessive emphasis on mechanical learning.

   

  *Besides the participation of Prof. Christoph Benzmüller, of FU Berlin, with his expertise, the roundtable also had presentations by Professor Astrid Ulloa (National University of Colombia) and Professor Laymert Santos (Campinas State University). Comments were the responsibility of Dr. Maya Manzi, Mecila Post-doctoral researcher, and the discussion was moderated by Professor Barbara Göbel (Ibero-Amerikanisches Institut & FU Berlin).

   

  By Ana Paula Katz Calegari