The paper aims at presenting some issues that arose when dealing with societal and ethical implications of AI since the seventies in which the author was involved. It is a narrative on how the understanding of AI dealt firstly with the question whether machines can think. With the rise of the internet in the nineties the perception of AI turned, secondly, into an issue of what AI as distributed intelligence means with an impact at all levels of social life no less that at basic ethical issues of daily life. In a breath-taking use of AI for all kinds of societal goals and contexts, the awareness grew, thirdly, that all natural and artificial things might be digitally connected with each other and to human agents. In the conclusion some challenges relating to the development and use of artificial intelligences are mentioned as well as results of recent research done in academia, scientific associations and political bodies concerning the possibilities for good life with and without artificial intelligences.
What does Artificial Intelligence (AI) mean in a broad historical perspective? This is a question that has not only sociological implications but addresses the basic understanding of technology as not being purely instrumental but shaping the relation between man and world. AI is the spirit of our time that conditions but does not determine knowledge and action. The answer to this question is a long and complex analysis going back to the roots not only of Western philosophy but also to other philosophical traditions. My aim in this paper is to recall some facts and discuss some arguments related to societal and ethical implications of AI particularly since the seventies in which I was involved.
My narrative about AI deals firstly with the question originated from cybernetics ―whether a machine can think― and I do this with reference to authors such as Alan Turing, Norbert Wiener, Joseph Weizenbaum, and Hubert Dreyfus. With the rise of the Internet in the nineties the perception of AI turned, secondly, into an issue of what AI as distributed intelligence means with an impact at all levels of social life. This broad societal challenge was called cyberspace; it was commonly perceived as a kind of separate sphere from the real world. This dualism soon became untenable. In a breath-taking development of digital technology for all kinds of societal goals and contexts, the awareness grew, thirdly, that all natural and artificial things might be digitally connected with each other as well as to human agents into what is being called the Internet of Things. The interpretation of AI changed from the original question whether machines can think into the one of what natural and artificial things are when they become intelligent.
I. Can machines think?
At every stage of technique since Daedalus or Hero of Alexandria, the ability of the artificer to produce a working simulacrum of a living organism has always intrigued people. This desire to produce and to study automata has always been expressed in terms of the living technique of the age. In the days of magic, we have the bizarre and sinister concept of the Golem, that figure of clay into which the Rabbi of Prague breathed life with the blasphemy of the Ineffable Name of God. In the time of Newton, the automaton becomes the clockwork music box, with the little effigies pirouetting stiffly on top. In the nineteenth century, the automaton is a glorified heat engine, burning some combustible fuel instead of the glycogen of the human muscles. Finally, the present automaton opens doors by means of photocells, or points guns to the place at which a radar beam picks up an airplane, or computes the solution of a differential equation. (Wiener 1965, 39-40)
enlarge this history with regard to literature (Karol Capek: R.U.R.
Universal Robots 1921, Stanislaw Lem: Golem XIV, 1981; Isaac Asimov's
Laws of robotics became famous through the collection I, Robot, 1950)
particularly to film (Fritz Lang: Metropolis 1927; Stanley Kubrick:
Space Odyssey, 1968; Aaron Liptsadt:
Android, 1982; Ridley Scott: Blade Runner 1982; Gene Roddenberry: Star
1987-1994, Albert Pyn: Cyborg, 1989;
Steven Spielberg: A.I. Artificial Intelligence 2001; Alex Proyas: I.
2004). The term artificial intelligence was first used in a scientific
in a workshop at
The word cybernetics is of Greek origin. A cybernetes is the pilot of a ship facing the insecurity of starting or not a travel in view of the weather, the sea, the robustness of the ship, the support of the crew. The Greeks called metis savvy intelligence useful for any kind of risky endeavours. Metis has to do with skills, prudence, wisdom, cunning, and trickery (Detienne and Vernant 1974). In a foundational text of Western thought, Aristotle writes in his Politics that in order to live well (eu zen) lifeless (apsycha) and living (empsycha) instruments (organon) are needed for the administration of the household (oikia). The rudder is such a lifeless instrument for the pilot of a ship, while a look-out man is a living one. Similarly, a slave (doulos) is a living possession which takes the place of all other instruments. He writes:
If every instrument could accomplish its own work, following a command or anticipating it, like the statues of Daedalus, or the tripods of Hephaestus, which, says the poet, of their own accord (automatous) entered the assembly of the Gods; if, in like manner, the shuttle would weave and the plectrum touch the lyre without a hand to guide them, the master builder (architektosin) would not need servants (hypereton), nor masters (despotais) slaves (doulon). (Aristotle, Politics, 1253 b 25-39, revised English translation, RC).
Aristotle addresses ironically a mythical society as a utopia where work is not based on the use of slaves but on lifeless intelligent automata. Karl Marx quotes this text in Das Kapital by saying that neither Aristotle, "the greatest thinker of antiquity," nor other thinkers could comprehend
[t]he economic paradox, that the most powerful instrument for shortening labour-time, becomes the most unfailing means for placing every moment of the labourer’s time and that of his family, at the disposal of the capitalist for the purpose of expanding the value of his capital. (Marx 1867, 335; Engl. transl. 2015, 278)
The use of machines based on steam-power, electricity or digital technology creates new forms of the division of labour under, according to Marx but also to Norbert Wiener, new slave-like conditions. Wiener wrote in 1950:
Let us remember that the automatic machine,
whatever we think of any
feelings it may have or may not have, is the precise economic
slave labor. Any labor which competes with slave labor must
the economic conditions of slave labor. It
is completely clear that this will produce an unemployment situation,
in comparison with which the present recession and even the depression
thirties will seem a pleasant joke. (Wiener 1989, 162)
It was Joseph Weizenbaum who in his book Computer Power and Human Reason (Weizenbaum 1976) also raised fundamental ethical issues of computer technology. The book was published ten years after his famous ELIZA — A Computer Program for the Study of Natural Language Communication between Man and Machine (Weizenbaum 1966). Weizenbaum opus magnum was a result of self-critical thinking. Herbert Simon published his The Sciences of the Artificial in 1969. In 1972 Hubert L. Dreyfus published the influential book What Computers Can't Do. The Limits of Artificial Intelligence (Dreyfus 1972). Other studies dealing with AI followed, such as Margaret Boden Artificial Intelligence and Natural Man (1977), Aaron Sloman The Computer Revolution in Philosophy (1978), Daniel C. Dennett Brainstorms (1978); Pamela McCorduck: Machines Who Think (1979); Don Ihde Technics and Praxis. A Philosophy of Technology (1979; John R. Searle Minds, Brains, and Programs (1980); Deborah G. Johnson Computer Ethics (1985); Terry Winograd & Fernando Flores Understanding Computers and Cognition (1986); P. S. Churchland Neurophilosophy (1986); Margaret Boden (ed.): The Philosophy of Artificial Intelligence (1990)
A number of these scholars were deeply influenced, as I was, by the traditions of hermeneutics and phenomenology. Those who stand out include Hubert Dreyfus ― I had the privilege to meet him in 1992 (Capurro 2018a) ― and Terry Winograd and Fernando Flores. Winograd and Flores attracted my attention on AI in the eighties and early nineties after I published my post-doctoral thesis Hermeneutics of Specialized Information (Capurro 1986). The work analyzed the relationships between human understanding and interaction with computer-based information storage and retrieval.
In 1987 I was invited by German philosophers Hans Lenk and Günter Ropohl to write a contribution on the emerging field of computer ethics for a reader on Technology and Ethics published by Reclam well known for their little yellow paperback books. The book included contributions by Theodor W. Adorno, Hans Jonas, Kenneth D. Alpern and Alois Huning dealing particularly with ethical issues of engineering (Lenk & Ropohl 1987). Later on, I wrote two papers on Joseph Weizenbaum whom I had the privilege to meet several times (Capurro 1987, 1998).
In 1987 I made a short presentation on AI at the 14th German Congress of Philosophy. The argument was, following Dreyfus, that while AI is based on explicit formalized rules, human understanding is incarnate in a body, sharing with others a common world and related to a situation that it transcends (Capurro 1987a). Winograd and Flores addressed this difference with regard to the design of computer systems that I analyzed in a paper published in the Informatik-Spektrum, the German journal of Computer Science (Capurro 1987b, 1992; see also: Capurro 1988, 1988a, 1989, 1990).
In 1988 I
participated at the 18th World Congress of Philosophy held in
It was the Jesuit theologian Karl Rahner (1904-1984) who in his book Geist in Welt (Spirit in the World) (Rahner 1957) opened my eyes about the relevance of angelology when dealing with the nature of human knowledge. According to Rahner, who develops his argument based on a detailed analysis of Thomas Aquinas quaestio 89 of the Summa Theologiae, human knowing cannot be compared with God's knowledge as it is completely different to ours. A comparison as thought experiment, with angel's knowing lets the differentia specifica of human knowledge shine forth. Being both creatures, humans and angels have knowledge in common as tertium comparationis. But humans are incarnated spirits that need to go back to sensory experience (conversio ad phantasmata) after the process of abstraction (abstractio) of the forms, which is not the case with angels. The view of divine perennial substances (aidiai ousiai) separated from the finite material natural processes of becoming and decay ―the Latin terminus technicus being intelligentiae separatae― goes back to Aristotle's Metaphysics Book Lamda (Aristotle 1973). Aquinas' epistemological reflection on angels and humans is based on Latin translations of Islamic philosophers such as Avicena, who themselves translated and commented Greek classics. In this context of Greek-Persian-Arabic-Latin tradition, I first discovered the Greek origin (eidos, idea, typos, morphe) and the Latin roots of the term information (informatio) as well as its relation to the concepts of messenger and message (angelos/angelia) (Capurro 1978, 2018; Capurro & Holgate 2011). Aquinas' angelology has its source in the Bible, according to which angels are immortal but not eternal creatures. In many cases they are God's messengers and not just, as in the case of Aristotle, perennial star movers (Capurro 1993, 2017).
Living in a secular and technological age, the idea of creating artificial intelligences that would even supersede the human one can be considered in some way in parallel to ancient and medieval thinking about divine and human intelligence, the relata being now human (natural) and artificial (digital). Artificial intelligences would not only enhance but eventually supersede human intelligence as the debate on transhumanism and singularity shows (Eden et al. 2012) I developed in the early nineties a critique of authors like Hans Moravec (1988) and Raymond Kurzweil (1999) on what I called cybergnosis (Capurro & Pingel 2002) an issue that raised my attention in the early nineties as related to the analogy between angels and computers (Capurro 1993; 1995, 78-96). According to Blaise Pascal: "L'homme n'est ni ange ni bête, et le malheur veut que qui veut faire l'ange fait la bête." (Pascal 1977, 572) The English translation "Man is neither angel nor beast, and whoever wants to act as an angel, acts the beast" does not reflect the double meaning of the verb "faire" ― to act but also to make ―, although this second meaning is not the one addressed by Pascal. Many years later, in 2010, I was invited to participate at an international conference on Information Ethics: Future of Humanities, organized by the Oxford Uehiro Centre for Practical Ethics, the Uehiro Foundation on Ethics and Education, and the Carnegie Council for Ethics in International Affairs, that was held at St Cross College, Oxford. I presented a paper with the title Beyond Humanisms. My argument was that Western humanisms rest on a fixation of the humanum. They are metaphysical, although they might radically differ from each other. I addressed the debate on trans- and posthumanism as follows:
The question about the humanness of the human and its "beyond" is not any more concerned with the relationship between the human and the divine as was the case with the classical humanisms in Antiquity, Renaissance and Reformation, nor with the self-introspection of the subject as in Modernity, but with the hybridization of the human, particularly through the digital medium as well as through the possibilities to change the biological substrate of the human species. A common buzz-word for these issues is "human enhancement." (Capurro 2012b, 49-50).
difference between strong and weak AI was one of the main issues
the nineties (Capurro 1993). It dealt with the question how far
can be separated from its biological substrate being (or
product of programming in the digital medium. The strong dualistic
more and more problematic considering that matter matters, that is to
natural intelligence is intrinsically related to its embodiment and
bottom-up procedure must take the issue of the medium seriously or
consider that what is crucial with the concept of artificial
not the asymptotic and unachievable approach to human intelligence but
difference created when working with another medium. When dealing with
artificial intelligence(s) a key issue is to clarify the concept of
artificiality. It was the Italian sociologist Massimo Negrotti who
eyes on this matter (Negrotti 1999, 2002, 2012; Capurro 1995a). The
between hard- und software, underlying the strong AI thesis finds its
counterpart in the metaphysical dualism between human and angelic
intelligences. This dualism is portrayed by Lewis Carroll in the
II. Distributed Intelligence
ethical and legal issues of AI that were mainly object of academic
exploded in a global context that made manifest different research
backgrounds and every-day lifeworlds of different societies. As the
took root so did concerns for an ever-growing schism between those with
those without access to its benefits. This was soon to be known as the digital
divide. At the same time, it became evident that this was not only
technical issue. Basic questions related to privacy and
were at stake. The World Summit on the Information Society (WSIS)
the United Nations that took place 2003 in
issues of AI were discussed at the beginning of the new century in two
projects in which I participated, namely ETHICBOTS (2005-2008) and
(2009-2011). The ETHICBOTS (Emerging Technoethics of Human
Communication, Bionic and Robotic Systems) project took place under
leadership of the Italian philosopher Guglielmo Tamburrini, University
1. The triaging categories of imminence, novelty, and social pervasiveness to assess the urgency of and the need for addressing techno-ethical issues. 2. A variety of ethical approaches and perspectives to represent the ethical richness of the European culture and tradition. (Capurro, Tamburrini, Weber 2008, 14)
of the project included a paper on Ethical Regulations on Robotics
(Nagenborg, Capurro, Weber, Pingel 2008) as well a book on Ethics
Robotics (Capurro & Nagenborg 2009). In a contribution to
the workshop L'uomo e la macchina. Passato e presente (Pisa
organized by the Philosophy Department of the
The ETICA (Ethical Issues of Emerging ICT Applications) project (2009-2011), under the leadership of the German philosopher Bernd Carsten Stahl, dealt with the following technologies: affective computing, ambient intelligence, artificial intelligence, bioelectronics, cloud computing, future internet, human-machine symbiosis, neuroelectronics, quantum computing, robotics, virtual/augmented reality. In the Ethical Evaluation by Michael Nagenborg and myself we summarized the ethical issues of AI as follows:
1. Human Dignity: The visions of ―artificial persons or ―artificial (moral) agents with corresponding rights are to be seen as being in contrast to the emphasis given to human rights in the European Union. This might be even stronger the case with anthropomorphic robots.
2. Autonomy and responsibility: The question of 'machine autonomy‘ does give rise to questions about human autonomy and responsibility.
3. Privacy: AI is one of the major building blocks of surveillance society.
4. Cultural Diversity: Artificial moral agents with a strong bias towards a certain cultural identity might be in contrast to a pluralistic society.
5. Inclusion: AI might contribute in making ICT more accessible to many people, but it might also foster the digital divide.
6. Access to the labour market: AI systems are likely to replace humans in certain contexts
7. Precautionary Principle: The precautionary principle might be invoked with regard to military applications of AI.
8. Principle of Transparency: The potential (bi-directional) dual use of AI systems calls for paying attention to the funding and future use of R&D in the field.
9. Likelihood of Ethical Issues: High. (Nagenborg & Capurro 2012, 20-21)
In the recommendations we stated
[t]he current research on Computer and Information Ethics is very much human-centred, which means that there is little to none research on animals or environmental issues. Therefore, we would like to encourage our colleagues to take some inspirations from the Ethics of the European Institutions and to overcome the bias towards humans. (Nagenborg & Capurro 2012, 75)
In the Annex
to this deliverable Lisa Stengel and Michael Nagenborg analyzed the
how does a technology become an ethical issue at the level of the EU,
as in the
case of the work done by the European Group on Ethics in Science
Technologies (EGE) to the European Commission as well as by the
Ethics Committees. They remarked: "Since the European Community moved
a mere economic community to a political
According to The Charter of Fundamental Rights of the European Union such "community of values" consist of human dignity, freedom, democracy, equality, the rule of law and the respect for human rights (Nagenborg & Stengel 2012, 1-2). The EGE issued an Opinion in 2005 on Ethical Aspects of ICT Implants in the Human Body (EGE 2005) of which Stefano Rodotà and myself were the rapporteurs that raised questions related to AI from the perspective of the European "community of values." We summarized the issues as follows:
“We shall not lay hand upon thee”. This was the promise made in the Magna Charta – to respect the body in its entirety: Habeas Corpus. This promise has survived technological developments. Each intervention on the body, each processing operation concerning individual data is to be regarded as related to the body as a whole, to an individual that has to be respected in its physical and mental integrity. This is a new all-round concept of individual, and its translation into the real world entails the right to full respect for a body that is nowadays both physical and electronic. In this new world, data protection fulfils the task of ensuring the “habeas data” required by the changed circumstances – and thereby becomes an inalienable component of civilisation, as has been the history for habeas corpus.
At the same time, this is a permanently unfinished body. It can be manipulated to restore functions that either were lost or were never known – only think of maiming, blindness, deafness; or, it can be stretched beyond its anthropological normality by enhancing its functions and/or adding new functions – again, for the sake of the person’s welfare and/or social competitiveness, as in the case of enhanced sports skills or intelligence prostheses. We have to contend with both repairing and capacity enhancing technologies, the multiplication of body-friendly technologies that can expand and modify the concept of body care and herald the coming of 'cyborgs' – of the posthuman body. “In our societies, the body tends to become a raw material that can be modelled according to environmental circumstances”. The possibilities of customised configuration undoubtedly increase, and so do the opportunities for political measures aimed at controlling the body by means of technology.
The downright reduction of our body to a device does not only enhance the trend ―already pointed out― towards turning it increasingly into a tool to allow continuous surveillance of 29 individuals. Indeed, individuals are dispossessed of their own bodies and thereby of their own autonomy. The body ends up being under others’ control. What can a person expect after being dispossessed of his or her own body? (EGE 2005, 39-30)
Stefano Rodotà (1933-2017), a famous Italian jurist and politician, published the opus magnum Treatise of Biolaw (Trattato di Biodiritto) edited together with Paolo Zatti, the first volume being Field and Sources of Biolaw (Ambito e Fonti del Biodiritto) edited by Mariachiaria Tallacchini and himself. In a comprehensive contribution on "the new habeas corpus" Rodotà who, as far as I know, first coined the concept of "habeas data" used in the 2005 EGE Opinion, analyzes key ethical and legal issues of the digitization of the human body (Rodotà 2010, 169-230). According to Rodotà, the basic right to informational self-determination, overcomes the dualism between habeas corpus, dealing with the protection of the physical body, and habeas data, dealing with the protection of the electronic one. There are not different subjects to be protected but a common object, namely "the person in its different configurations, conditioned little by little in its relation with the technologies that are not only the electronic ones." (Rodotà 2010, 229, my translation). The EGE issued two Opinions dealing with Ethics of Information and Communication Technologies (EGE 2012) and with Ethics of Security and Surveillance Technologies (EGE 2014).
development and use of AI devices, particularly robots, became
diversified due to the impact of the Internet in all areas of social
question in 1950 "can machines can think?" turn more and more on the
intrinsic relation of the practical issue concerning their "intelligent
behavior," Turing's formulation from 1948. Building robots and the
reflection upon them becomes a social or moral issue, that is
to say, it
concerns contexts of application with specific
values, customs and rules of behavior (Latin mores) and a
reflection thereupon. In 2009 the Center for Cybernetics Research
The real intercultural ethical challenge in Japan is, I think, to ponder how robots become part of Japanese interplay between Japanese minds, which differs from the interplay in the "Far West," ― particularly as it is based on the Buddhist tradition of 'self-lessness' or Mu ― sharing a common Ba [place]" (Nakada & Capurro 2013, 14; see also Nakada, Capurro, Sato 2017; Capurro 2017a; Tzafestas 2016, 155-175).
The nature of the self is indeed a key issue when dealing with the question about the interaction between artificial intelligences and human beings, being the case that artificial intelligences might mimic a self but, in fact, they are not (so far) a who but a what. In an interdisciplinary project organized by ACATECH (German Academy for Science and Engineering), on the question of privacy and trust on the internet, the ethics group, composed of the Australian philosopher Michael Eldred, the German lawyer Daniel Nagel and myself, developed a view of privacy and trust based on this difference. In the introduction to the ethics chapter of the final report we stated:
The concept of privacy cannot be adequately determined without its counterpart, publicness. Privacy and publicness are not properties of things, data or persons, but rather ascriptions dependent upon the specific social and cultural context. These ascriptions relate to what a person or a self (it may also be several selves) divulges about him- or herself. A self, in turn, is not a worldless, isolated subject, but a human being who is and understands herself always already connected with others in a shared world. The possibility of hiding, of displaying or showing oneself off as who one is, no matter in what way and context and to what purpose, is in this sense, as far as we know, peculiar to human beings, but precisely not as the property of a subject, but rather as a form of the interplay of a human being's life as shared with others. (Capurro, Eldred, Nagel 2012, 64)
In a special
section of this chapter I analyzed intercultural aspects of digitally
whoness, privacy and freedom in the Far East, Latin America and
III. Natural and Artificial Intelligences
Since then, robots have become increasingly socially relevant beyond industry applications and an entirely new conception, the Internet of Things, has in conjunction with RFID technology become a window into future possibilities. The term, Internet of Things, was coined by Kevin Ashton in 1999. Asthon wrote in 2009:
We're physical, and so is our environment. Our economy, society and survival aren't based on ideas or information—they're based on things. You can't eat bits, burn them to stay warm or put them in your gas tank. Ideas and information are important, but things matter much more. [...] We need to empower computers with their own means of gathering information, so they can see, hear and smell the world for themselves, in all its random glory. RFID and sensor technology enable computers to observe, identify and understand the world—without the limitations of human-entered data. Ten years on, we've made a lot of progress, but we in the RFID community need to understand what's so important about what our technology does, and keep advocating for it. It's not just a "bar code on steroids" or a way to speed up toll roads, and we must never allow our vision to shrink to that scale. The Internet of Things has the potential to change the world, just as the Internet did. Maybe even more so. (Ashton 2009)
The Internet was once an idea and is now a thing. The Internet of Things is an idea in the process of becoming a thing. It goes beyond the traditional view of artificial intelligent things called robots, envisaging the transformation of all kinds of natural and artificial things becoming digitally networked, although this might not be necessary the case. It is, I believe, the powerful IT industry with its insatiable hunger for digital data from users viewed as customers which is strongly behind the idea and the reality of the Internet of Things. Intelligent things might be also designed and used as stand-alone but this is bad news for IT giants. This is also true of humans as far as they might design their lives as being always online, what is being called onlife. Things and humans would eventually turn just into a bunch of digital data. A "who" might be confused with a "what" and a "what" might look like a "who." The difference between living online and offline does not mean a kind of Platonic dualism of separate worlds, but concerns the freedom of choice, life design and the protection of privacy. George Orwell's Animal Farm (Orwell 1989) might turn into a digital farm. Without ideas nothing would change in the human world and, what is even more relevant, ideas make possible for us, humans, to reflect on the foreseen and unforeseen changes that things bring about for as well as for non-human living and non-living, natural and artificial beings with whom we share a common world.
What are things such as the Internet or the Internet of Things? They are, prima facie, just tools. What is a tool? In Heidegger's famous tool analysis in Being and Time he coined the terms "readiness-to-hand" and "presence-at-hand." When tools break down, that is to say, when they lose there readiness-to-hand, the worldly context or structure of references to which they implicitly belong, becomes manifest (Heidegger 1987, 102ff; Capurro 1992). In the introduction to Understanding Computers and Cognition written sixty years after Heidegger's work, Winograd and Flores write how: "[...] in designing tools we are designing ways of being" (Winograd & Flores 1986, xi). It is not due to the Internet that things are embedded in semantic and pragmatic networks but the other way round. It is because they, that is to say, we are from scratch embedded in such networks that we are able to design a digital network and interact with them and between ourselves. This means a paradigm change with regard to the way modernity conceived things as objects in the so-called "outside world" to be the correlate to an encapsulated subject. Not only humans and natural things but also artificial things become autonomous and networked agents in the digital age. But what does autonomy and action mean in each case? In a contribution to the panel on Autonomic Computing, Human Identity and Legal Subjectivity hosted by Mireille Hildebrandt and Antoinette Rouvroy at the International Conference: Computers, Privacy & Data Protection: Data Protection in a Profiled World that took place in Brussels in 2009 I wrote:
At today’s early stage of these breath-taking developments, it is difficult to give a typology of the new kinds of digital and living agents and the theoretical and practical challenges arising from them. From a broad perspective, these challenges are related, on the one hand, to all kinds of robots, starting with the so-called softbots (digital robots) as well as to all kinds of physical robots – including the (still speculative) nanobots based on nanotechnology – with different degree of complexity, including all forms of imitation of human and non-human living beings (bionics). On the other hand, there are the possibilities arising from the hybridization between ICT with non-human as well as with human agents, for instance. ICT, or other technologies, can become part of living organisms, for instance as implants (EGE 2005), or vice versa. In this case, humans become (or have already become) "cyborgs" (Hayles 1999). Finally, synthetic biology allows the artificial construction of new as well as the genetic modification of living beings (EGE 2009; Karafyllis 2003). (Capurro 2012a, 484)
The pervasive use of AI raises the question of the very basic understanding of technology as not being purely instrumental but shaping the relation between man and world. It belongs to what I call digital ontology, that is to say, the interpretation of the being of beings as well as of being itself from a digital perspective as a possible one. This ontological perspective might turn into a metaphysical world view or, politically speaking, into an ideology in case it becomes dogmatic, immunizing itself from critique (Capurro 2006, 2008, 2017c).
The Finish information security researcher Kimmo Halunen recently wrote a contribution with the title "Even artificial intelligences can be vulnerable, and there are no perfect artificial intelligence applications" (Halunen 2018). I asked him if he was the first one to use the plural noun "artificial intelligences" but he could not clarify the issue. In any case, the use of the plural noun might help to demystify the big noun AI by paying attention to a diversity of "artificial intelligence applications" making a difference with regard to other kinds of natural or artificial ones. Halunen writes:
Artificial intelligence has its own special characteristics that also make other kinds of attacks against these systems possible. Because an artificial intelligence usually attempts some kind of identification and then makes decisions based on it, the attacker may want to trick the artificial intelligence. This problem has been encountered in the fields of pattern and facial recognition in particular. Last year, it was published that Google’s artificial intelligence algorithm was tricked into classifying a turtle as a rifle. As for facial recognition, makeup and hairstyles that fool facial recognition algorithms have been developed. Of course, people also make mistakes in identifying objects or faces, but the methods used for identification by an artificial intelligence are very different. This means that the errors made by an artificial intelligence seem bizarre to humans, because even small children can tell a turtle from a rifle, and these camouflage methods do not work against people. In an automated environment, in which artificial intelligence makes the decisions, such deceptions can be successful and may help the attacker. (Halunen 2018)
artificial intelligences? Energy and human needs, beliefs and desires
in digital algorithms (Capurro 2019). It is not primarily a question
machines can think or how far they can be like human
even better ― other machines and living beings supersede humans
regards ― but on how we might be able to live with or without them in
contexts in the life-world. Artificial intelligences or, for that
computer programs can break down as Winograd and
contribution to the international conference: Artificial
& Regulation, organized by LUISS (Libera Università
Studi Sociali Guido Carli) held in
Algorithms are implicitly or explicitly designed within the framework of social customs. They are embedded in cultures from scratch. According to the phenomenologist Lucas Introna, creators and users are "impressed" by algorithms (Introna 2016). The "impressionable subject," however, is not the modern subject detached from the so-called outside world, but a plurality of selves sharing a common world that is algorithmically intertwined. What is ethically at stake when dealing with algorithms becomes part of human mores? What is the nature of this entanglement between human mores and algorithms? To what extent can it be said that algorithms are, in fact, cultural? Who is responsible for the decisions taken by algorithms? To what extent is this anthropomorphic view on algorithms legitimate in order to understand what algorithms are? These are some foundational questions when dealing with the ethics of algorithms that is in an incipient state (Mittelstadt et al. 2016). [...] The present casting of ourselves as homo digitalis (Capurro 2017) opens the possibility of reifying ourselves algorithmically. The main ethical challenge for the inrolling digital age consists in unveiling the ethical difference, particularly when addressing the nature of algorithms and their ethical and legal regulation. (Capurro 2019, forthcoming)
on driverless cars sometimes obfuscates basic questions on mobility
societies and individuals in the 21st century. At least some parts of
seem to be interested in these issues though. I received an invitation
from the Verband der Automobilindustrie (VDA) (German
Association of the
Automobile Industry) to a dialogue with the CEO of Continental AG,
Elmar Degenhart. The meeting took place in
Last but not least, I would like to mention the current issue of the International Review of Information Ethics dealing with "Ethical Issues of Networked Toys," the guest editors being Juliet Lodge and Daniel Nagel. In their introductory remarks they write:
Networked toys - Artificial guardians for little princesses or demonic plastic princes? Networked toys dominate the shelves in toy stores at a time when neither their real benefits nor their potentially latent dangers have been fully explored. Do hyper-connected toys transform the relationship between adults, the child and its environment? Do they shape their minds and predispose them to seek convenience and speedy responses rather than rely on their own autonomous capacities for critical thought?
Questions such as who really is in control arise, both of the toys ―parents, third parties or even the toddlers themselves― and of data (including biometrics) that might be collected for unclear purposes and opaque destinations. For what specific or linkable purpose and above all where and to whom is data transmitted? What ethical considerations should be addressed?
Is there an actual benefit for the children themselves? Do hyper connected devices and robo-toys teach them how to handle technology or does it erode their capacity for autonomous reflection as speed and convenience are prioritised in their on-line and off-line worlds? Do such toys presage fundamental transformation of childhood and the imagined and physical worlds? (Lodge & Nagel 2018)
Conclusion: Enlightening the Digital Enlightenment
What is the task of Information ethics in the era of artificial intelligences? Answer: enlightening the digital enlightenment which follows the path of thought of the philosophers Max Horkheimer and Theodor W. Adorno in their influential collection of essays published in revised edition in 1947 with the title Dialectics of Enlightenment (Dialektik der Aufklärung) (Horkheimer & Adorno 1975). A main insight of this book is the ambivalence of the project(s) of enlightenment coming from the social revolutions of the nineteenth century but going back to the dialectics between mythology and science that characterizes European Enlightenment particularly in the eighteenth century. Enlightenment must take care of this ambivalence that might revert digital enlightenment into digital mythology. This narrative shows the changing meanings of the concept of artificial intelligence(s) since the middle of the last century depends both on the state-of-the-art of digital technology as well as of the different contexts in which it has been used. Looking back to my personal experiences since the early seventies and the changing academic debates in the years that followed, I dare no forecast beyond what appears today as challenges in the near future.
The task of
taming the digital chaos through different kinds of national and
regulations is still very much in the early stages and is dependent on
awareness of these issues take root across the globe. Enlightened
addresses several problems such as ecological issues, sustainability,
state regulation, fake news, cyber wars, digital capitalism, digital
colonialism, social justice, surveillance society, digital addiction,
future of work, and who we are as cybercitizens in the digital age.
and Lynette Schultz organized an Information Ethics Roundtable at the
Who should take care of the enlightenment of the digital enlightenment? Answer: universities, research institutions, scientific associations, governments and the media. As a paramount example of a scientific association leading in the field of enlightening the digital enlightenment I would like to mention the Institution of Electrical and Electronics Engineers Standards Association (IEEE). It brought about the Global Initiative of Autonomous and Intelligent Systems, under the leadership of managing director Konstantinos Karachalios. John Havens took care as Executive Director of the Ethical Considerations in Artificial Intelligence and Autonomous Systems. Jared Bielby managed the Committee for Classical Ethics. The final report represents the collective input of several hundred participants from six continents who are thought to be leaders from academia, industry, civil society, policy and government (IEEE 2016). I quote the introduction in extenso:
The task of the Committee for Classical Ethics in Autonomous and Intelligent Systems is to apply classical ethics methodologies to considerations of algorithmic design in autonomous and intelligent systems (A/IS) where machine learning may or may not reflect ethical outcomes that mimic human decision-making. To meet this goal, the Committee has drawn from classical ethics theories as well as from the disciplines of machine ethics, information ethics, and technology ethics. As direct human control over tools becomes, on one hand, further removed, but on the other hand, more influential than ever through the precise and deliberate design of algorithms in self-sustained digital systems, creators of autonomous systems must ask themselves how cultural and ethical presumptions bias artificially intelligent creations, and how these created systems will respond based on such design. By drawing from over two thousand years’ worth of classical ethics traditions, the Classical Ethics in Autonomous and Intelligent Systems Committee will explore established ethics systems, addressing both scientific and religious approaches, including secular philosophical traditions such as utilitarianism, virtue ethics, deontological ethics and religious and culture-based ethical systems arising from Buddhism, Confucianism, African Ubuntu traditions, and Japanese Shinto influences toward an address of human morality in the digital age. In doing so the Committee will critique assumptions around concepts such as good and evil, right and wrong, virtue and vice and attempt to carry these inquiries into artificial systems decision-making processes. Through reviewing the philosophical foundations that define autonomy and ontology, the Committee will address the potential for autonomous capacity of artificially intelligent systems, posing questions of morality in amoral systems, and asking whether decisions made by amoral systems can have moral consequences. Ultimately, it will address notions of responsibility and accountability for the decisions made by autonomous systems and other artificially intelligent technologies. (IEEE 2016)
political level I highlight and support the recent activities of the
Union, particularly the Communication from the Commission to the
Parliament, the European Council, the Council, the European Economic
Committee and the Committee of the Regions: Artificial Intelligence for
Aristotle (1973). Metaphysica. Ed.
Aristotle (1950a). Physica. Ed. W.
Ashton, K. (2009). The 'Internet of
Things' Thing. In the real world, things matter more than ideas. In:
Boden, M. A. (1977). Artificial
Intelligence and Natural
Boden, M. A. (ed.) (1990). The
Philosophy of Artificial Intelligence.
R. (1978). Information. Ein Beitrag zur etymologischen und
Begründung des Informationsbegriffs. München: Saur.
R. (1986). Hermeneutik der Fachinformation (Hermeneutics of Scientific
Information). Freiburg/München: Alber.
Capurro, R. (1987). Zur
Computerethik. Ethische Fragen der Informationsgesellschaft
Ethics. Ethical Questions of the Information Society). In: H. Lenk, G.
(eds.): Technik und Ethik. Stuttgart: Reclam, 259-273.
Capurro, R. (1987a). Zur
Kritik der künstlichen Vernunft. Oder über die Einheit und
Intellekts (A Critique of Artificial Intelligence. On the Unity and
of the Intellekt). In: O.
Marquard, P. Probs (eds.): Proceedings of the 14th
German Congress of Philosophy,
R. (1987b). Die Informatik und das hermeneutische Forschungsprogramm
Science and the Hermeneutical Research Programm). In:
R. (1988). Die Inszenierung des Denkens. In: Mensch – Natur –
5, Heft 1, 1988, 18-26.
R. (1988a) Von der Künstlichen Intelligenz als einem
ästhetischen Phänomen. Eine
kritische Reflexion in Kantischer Absicht (Artificial Intelligence from
Pespective. A Kantian Critique)
R. (1989). Stellt die KI-Forschung den Kernbereich der Informatik dar? (Is AI Research the Core of Computer
Science?). In: J. Retti & K. Leidlmaier (eds.): 5. Österreichische
Artificial-Intelligence-Tagung: Igls/Tirol, 28.-31. März 1989,
1989, S. 415-421.
Capurro, R. (1989a). Der Kongress. Rafael
Capurro über den XVIII. Weltkongress für Philosophie
in Brighton, GB
(21.-27. August 1988) (The
Congress. Rafael Capurro on the 18th World Congress of Philosophy in
R. (1990). Ethik und Informatik (Etthics and Informatics) In:
R. (1992) Informatics and Hermeneutics. In: Ch. Floyd, H. Züllighoven, R. Budde, R. Keil-Slawik
(eds.): Software Development and Reality Construction. Berlin:
R. (1993). Ein Grinsen ohne Katze. Von der Vergleichbarkeit zwischen
'künstlicher Intelligenz' und 'getrennten Intelligenzen' (A Grin
without a Cat.
On the Comparison between AI and Separate Intelligences). In:
für philosophische Forschung, 47, 93-102.
R. (1995). Leben im Informationszeitalter (Living in the Information
Berlin: Akademie Verlag.
R. (1995a). On Artificiality. Working
paper published by IMES (Istituto Metodologico Economico Statistico)
for the Culture of the Artificial, Università di Urbino,
R. (1998). Die Verantwortbarkeit des Denkens. Künstliche
Intelligenz aus ethischer
Sicht (The Responsibility of Thinking. AI from an Ethical Perspective).
In: Forum für
interdisziplinäre Forschung 1,
Capurro, R. (2006). Towards an
Ontological Foundation of Information Ethics. In: Ethics and
Technology 8 (4) 175-186
Capurro, R. (2008). On Floridi's
Metaphysical Foundation of Information Ecology. In: Ethics and
Information Technology, 10 (2-3) 167-173.
Capurro, R. (2009). Ethics and
Robotics. In: R. Capurro, M.
Nagenborg (eds.). Ethics and Robotics. Heidelberg: Akademische
R. (2010). Digital Ethics. In: The
Capurro, R. (2010a). Ethics and
Public Policy in
Capurro, R. (2011). The Quest for
Roboethics: A Survey. In: T. Kimura, M. Nakada, K. Suzuki, Y. Sankai
Technical Reports. Special Issue on Roboethics. University of Tsukuba,
- CYB-2011-008, 39-59.
Capurro, R. (2012). Intercultural Aspects of Digitally Mediated
Whoness, Privacy and Freedom. In: R. Capurro,
R., M. Eldred, D. Nagel: ‘IT and Privacy from an Ethical Perspective:
Whoness: Identity, Privacy and Freedom in the Cyberworld. In: Johannes
(ed.) Internet Privacy - Eine multidisziplinäre
Multidisciplinary Analysis. Acatech Studie, Berlin, 113-122.
Capurro, R. (2012a). Towards a
Comparative Theory of Agents. In: AI
& Society, 27 (4), 479-488.
Capurro, R. (2012b). Beyond
Humanisms. In: T. Nishigaki, T. Takenouchi (eds.): Information Ethics.
Future of the Humanities,
Capurro, R. (2017). Engel, Menschen
und Computer. Zur Relevanz der thomistischen
Engellehre für die philosophischen Anthropologie. (Angels, Humans and Computers.
On the Relevance of Tomas Aquinas' Angelology for the Philosophical
Roboethics for a Robot Age. In:
M. Nakada, R. Capurro, K. Sato (eds.) (2017). Critical Review of Information Ethics and Roboethics
in East and
West. Master's and Doctoral Program in International and Advanced
Studies, Research Group for "Ethics and Technology in the Information
R. (2017b). Fahrer entlasten nicht ersetzen (Helping Drivers not
Them) In: Flotten management 1,
Capurro, R. (2017c). Homo
Digitalis. Beiträge zur Ontologie, Anthropologie und Ethik der
Technik. (Homo Digitalis.
Ontological, Anthropological and Ethical Contributions to Digital
Capurro, R. (2017d). Citizenship in
the Digital Age. In: T. Samek, L. Schultz (eds.): Information Ethics,
Globalization and Citizenship. Essays on Ideas to Praxis.
Robotic Natives. Leben mit Robotern im 21.
Jahrhundert. In: R. Capurro: Homo Digitalis. Beiträge
zur Ontologie, Anthropologie und Ethik der digitalen Technik.
Capurro, R. (2018). Apud Arabes.
Notes on Greek, Latin, Arabic, Persian, and Hebrew Roots of the Concept
Capurro, R. (2018a). A Long-Standing Encounter. In: AI
& Society, 1-2.
Issues of Humanoid-Human Interaction. In: A.
Goswami, P. Vadakkepat (eds.): Humanoid Robotics: A Reference.
Capurro, R., Eldred, M., Nagel, D.
(2012): ‘IT and Privacy from an Ethical Perspective: Digital Whoness:
Privacy and Freedom in the Cyberworld. In: J. Buchmann
Privacy – Eine multidisziplinäre Bestandsaufnahme. A
Analysis. Acatech Studie, Berlin, 63-142.
Capurro, R., Holgate, J. (eds.) (2011). Messages and Messengers. Angeletics as an Approach to the Phenomenology of Communication. München: Fink.
Carroll, L. (1965). The Annotated
Churchland, P.S. (1986).
Neurophilosophy: Toward a unified science of the mind-brain.
Copeland, B. J., Proudfoot,
D. (1999). Alan Turing's Forgotten
Ideas in Computer Science.
In: Scientific American, April, 99-103.
A.B., Haberbeck, R., Seetzen, J.
Wachsmuth, I. (eds.) (1992): Künstliche Intelligenz.
Verantwortbarkeit (Artificial Intelligence. Leading
Ideas and Responsibility) VDI Report 17, 1992
Capurro, R., Tamburrini, G., Weber
J. (2008). ETHICBOTS, D
5: Techno-ethical Case-Studies in Robotics, Bionics, and Related AI
Capurro, R., Nagenborg, M. (eds.) (2009). Ethics and Robotics. Heidelberg: Akademische Verlagsgesellschaft.
Dennett, D. C. (1978). Brainstorms:
Philosophical Essays on
Mind and Psychology.
Detienne, M., Vernant, J.-P. (1974). Les ruses de l'intelligence. La mètis des Grecs. (Cunning Intelligence in Greek Culture and Society). Paris: Flammarion.
Dreyfus, H.L. (1972). What
Computers Can't Do. The Limits
of Artificial Intelligence.
European Group on Ethics in Science
and New Technologies (EGE) (2005). Opinion No. 20: Ethical Aspects of
Implants in the Human Body.
European Group on Ethics in Science and
New Technologies (EGE) (2009). Opinion No. 25: Ethics of
European Group on Ethics in Science and
New Technologies (EGE) (2012). Opinion No. 26: Ethics of information
European Group on Ethics in Science and
New Technologies (EGE) (2014). Opinion No. 28: Ethics of Security and
Union (2018). Communication Artificial Intelligence for
Flores Morador, F. (2015). Encyclopedia
of Broken Technologies - The
Humanist as Engineer (October 23, 2015). Broken Technologies.
Humanist as Engineer. Ver 3.0, Lund University, Revised 2015.
A., Vadakkepat, P. (eds.) (2019) Humanoid Robotics: A Reference.
Halunen, K. (2018). Even artificial
intelligences can be vulnerable, and there are no perfect artificial
intelligence applications. VTT Technical Research Centre of
Heidegger, M. (1987). Being and Time. Oxford: Blackwell. Transl. J. Macquarrie, E. Robinson.
Horkheimer, M., Adorno, Th. W. (1975). Dialektik der Aufklärung (Dialectics of Enlightenment). Frankfurt am Main: Fischer.
IEEE Standards Association (2016). Ethically Aligned Design: A Vision for
Prioritizing Human Well-being with Autonomous and Intelligent Systems.
Don (1979). Technics and Praxis: A Philosophy of Technology.
Kurzweil, R. (1999). The Age of
Lenk, H., Ropohl, G. (eds.) (1987). Technik und Ethik (Technics and Ethics). Stuttgart: Reclam.
N.C. (2003). Versuch über den Menschen zwischen Artefakt und
Lebewesen. (Essay on Human Being between
and Living Being).
Lodge, J., Nagel, D. (eds.) (2018).
Ethical Issues of Networked Toys. In: International Review of
Ethics, Vol. 27.
P. (1979). Machines Who Think. A Personal Inquiry into the History and
Prospects of Artificial Intelligence.
McLuhan, M. (1962). The Gutenberg
Galaxy: the making of typographic man.
Marx, K. (1867). Das Kapital. Kritik der
politischen Ökonomie. Marx-Engels-Gesamtausgabe (MEGA), 1. Band.
Berlin: de Gruyter. http://telota.bbaw.de/mega/ (Engl. quote from
Capital. A Critique of Political Economy, Transl. S. Moore, E. Aveling
H. (1988). Mind Children:
The Future of Robot and Human Intelligence.
Nakada, M. (2011). Ethical and
Critical Analysis of the Meanings of 'Autonomy'
of Robot Technology in the West and
M., Capurro, R. (2013). An Intercultural Dialogue on Roboethics. In: M. Nakada, R.l Capurro (eds.):
The Quest for Information Ethics and Roboethics in East and West. Research
Report on trends in information ethics and roboethics in Japan and the
West. ReGIS (Research Group on the
Information Society) and ICIE (
Nakada, M., Capurro, R., Sato, K.
(eds.) (2017). Critical Review of Information Ethics and Roboethics in
West. Master's and Doctoral Program in International and Advanced
Studies, Research Group for "Ethics and Technology in the Information
Nagenborg, M., Capurro, R., Weber,
J., Pingel, Chr. (2008). Ethical Regulations on Robotics in
Nagenborg, M., Capurro, R. (2012).
Ethical Evaluation. Part of D.3.2 (Evaluation Report)
Negrotti, M. (1999). The Theory of
M. (2002). Naturoids. On the
Nature of the Artificial.
Negrotti, M. (2012). The Reality of
the Artificial. Nature, Technology and Naturoids.
G. (1987). Animal Farm. A
Pascal, B. (1977). Pensées.
Rahner, K. (1957). Geist in Welt (Spirit in the World). München: Kösel (2nd ed.)
Rodotà, S. (2010). Il nuovo Habeas Corpus: La persona costituzionalizzata e la sua autodeterminazione (The New Habeas Corpus: The New Constitutionalised Person and its self-determination). In: S. Rodotà, M. Tallachini (eds.): Ambito e fonti del biodiritto. Milano: Giuffrè, 169-230.
Samek, T., Schultz, L. (eds.):
Information Ethics, Globalization and Citizenship. Essays on Ideas to
Searle, J.R. (1980). Minds, Brains, and Programs. In: Behavioral and Brain Sciences 3 (3), 417-457.
Simon, H. (1969). The Sciences of
Sloman, A. (1978). The Computer Revolution
in Philosophy. Philosophy, Science and Models of Mind. The
Stahl, B.C. (2016). Ethics of
European Institutions as Normative Foundation of Responsible Research
Innovation in ICT. In: M. Kelly, J. Bielby (eds.): Information Cultures
Digital Age. A Festschrift in Honor of Rafael Capurro.
Stengel, L., Nagenborg, M. (2011).
European Ethics. How does a Technology become an Ethical Issue at the
the EU? ETICA, Deliverable 3.2.2
Turing, A. M. (1950). Computing Machinery and Intelligence. In: Mind, Vol. LIX, Issue 235, 433-460
Turing, A. M. (1948). Intelligent
Machinery, report written for the National Physical Laboratory.
Tzafestas, S.G. (2016). Roboethics. A Navigating Overview. Heidelberg: Springer
(Verein der Automobilindustrie) (2016a).
Dialogreihe Mobilität von
morgen. Gespräch mit dem Vorstandsvorsitzenden der
AG, Dr. Elmar Degenhart (Dialogue on Mobility Tomorrow with the
Continental AG, Dr. Elmar Degenhart).
Wallach, W., Allen, C. (2009): Moral
Machines: Teaching Robots Right from Wrong.
Weinberg, A. E. (1963). Science,
Government, and Information. A report of The President's Science
Weizenbaum, J. (1966). ELIZA — A Computer Program for the Study of Natural Language Communication between Man and Machine. In: Communications of the Association for Computing Machinery 9, 36-45.
Weizenbaum, J. (1976). Computer
Power and Human Reason: From Judgement To Calculation.
Wiener, N. (1965). Cybernetics: or
Control and Communication in the Animal and the Machine.
Wiener, N. (1989). The Human Use of
Human Beings. Cybernetics and Society.
Last update: June 30, 2020