• NATIONAL SALES MANAGER. WIRELESS...
• Calling All C++ & AI Game...
• C#ASP.NET Developers
• Technical Consultant
• CHANNEL ACCOUNT MANAGER $ 210K...

"Computer says no" - humans, ICT and ethics

Matthew Allen, Information Age

04/01/2008 03:26:42

The ethical dimensions of information technology are one of the enduring topics of conversation in academic and professional discussions about that technology's impacts, challenges and opportunities for our society.

The topic is standard fare in all computing and information systems courses of study, is a focus for professional development and the subject of considerable applied and pure research by academics.

What makes information technology and ethics fit together in this manner?

First, I would suggest that the rise and rise of computer-mediated activity in our society, whether in the commercial, governmental or private sphere, has made the profession of information technology much more significant, and with this change has come a much greater awareness amongst IT workers that they are, indeed, professional, in an equivalent manner to doctors, lawyers, accountants and the like.

In contemporary society, professions obtain that status by demonstrating their capacity for self-regulation and promotion of the key social role played by their members.

Thus, to be a profession necessarily involves members of the IT industry in committing to being, on the one hand, self-governing individuals and, on the other, as a collective which thinks seriously about the regulation of its own conduct. This dual process of self-regulation enables a group of people in society to claim for itself a privileged place of expertise, less subject to direct regulation by others and more trustworthy in its dealings with others.

Discussion of ethical behaviour by members of a profession underpins this outcome, for it both demonstrates that attention to self-regulation and also, through the effects of that discussion, aids in achieving the outcome sought. It should be noted, also, that recent and very public lapses in ethical - and indeed legal - conduct within the corporate world (think Enron and the like) have emphasised that professional ethics is a key element in the framework of governance of our free-market, largely privatised society.

Second, information technology is closely intertwined with considerations of ethics because the observable impacts of computerised, networked information processing by programmed machines have raised significant and very real questions about the kinds of activities and behaviours which constitute a "good" society.

Consider the following example. First, in Australia and other liberal democracies, it has been common practice for court records that reveal the criminal behaviour of guilty parties to be available to members of the public for many years prior to the emergence of the Internet. Such records were hard to access for the vast majority of people. While the Internet has not changed the formal status of those records from private to public it has, nevertheless, permitted organisations such as Crimenet (http://crimenet.com.au) to develop a system whereby those records become much more easily accessible and usable.

Thus, the effective status of the records has changed such that they are now far more public than before. This change in circumstances raises ethical questions about the consequences for society of making such information publicly available which previously were not significant (for more discussion, see Noble and Marchetti, 2005). It is arguable as to whether this practice is good or bad, in moral terms, but it is undeniable that it demonstrates the new ethical challenges which IT can bring.

Another example demonstrates how new challenges can also emerge, as well as "recasting" of older questions that bear on our ethical behaviour in relation to information. In this case consider the Internet's capacity for two-way, machine-level communication.

In 1999 Real Networks embedded within its software a unique identification code that would, unbeknown to its users, permit communication between users' and the company's computers to assemble a data profile on how the software was being used (principally, what audio or video files were being played).

The informatic nature of society's communications in the era of the Internet also shows how it can be hard to decide what ethical conduct might be, and also to commit to acting in an ethical manner (see Macavinta, 1999).

It is, therefore, easy to see why the information technology profession must (and does) regularly consider the ethical dimensions of its activities. But it is much less clear as to the directions we might take in answering the challenges that our society faces in this regard.

While there are numerous articles written in pursuit of these directions, considering either the educational steps that might be taken to promote ethical conduct, or discovering the real-world behaviours which might be deemed ethical or unethical (see for example, Walstrom, 2006 and Carbo and Amalgno, 2001), the answers probably lie in a consideration of more fundamental matters.

These concern the basis on which we might understand the complex interaction of information, computing and human endeavour and the intermingling of those three elements in our increasingly informatic world.

One key academic intervention in recent years to attempt to solve this problem - to provide the firm foundation on which to build a workable system of ethical conduct - comes from Luciano Floridi, currently Research Chair in Philosophy of Information at University of Hertfordshire and Fellow of the University of Oxford. Floridi has spent several years developing and proposing a new form of ethics, suitable to advanced societies whose very nature now depends on the way information and information technology works.

He terms it Information Ethics (see amongst other material, Floridi, 1999; 2002). It is designed to address very significant problems, such as "the preservation, dissemination, quality control and free flow of information; ...the enlargement of universal access... the respect for diversity and pluralism; [and] the overcoming of the digital divide (Floridi and Savulescu, 2006: 155).

Floridi's approach is relatively straightforward. He proposes that the information itself, in all its forms, should become the central point of departure for understanding ethical conduct. Information, he argues, exists within (indeed constitutes) an "infosphere" or a world that encompasses all of what we as humans do and are as information, and the products of that behaviour.

Within this world, humans are best understood as informational objects: that is, humans are themselves information, albeit incredibly complicated and taking on material form for the greater part of their being.

By characterising humans as information first, and human second, Floridi is then able to consider the relationship between other entities in that world which are, superficially entirely distinct from humans such as software, and hardware-software aggregations. Relatively easily, we can see that if humans are "information" then they exist on a continuum along with all other forms of information, and the fact that some of that information takes the form of machines, rather than flesh and blood, is a second-order issue.

Floridi's approach depends on a shift away from anthropocentrism that is not dissimilar to that proposed by Peter Singer. Singer's work, dating from at least his 1975 work, Animal Liberation, is also deeply concerned with the existence of humans within the biosphere, rather than as somehow apart from all other biological life on the planet).

He uses information as the central moment of his ethical system, rather than humans and his insights spring from that radical shift in thinking. In Floridi's infosphere, all objects - be they people, animals or pure informational representations - can be considered as part of the moral reckoning, due what Floridi terms "moral respect".

Floridi identifies that this move is both important for providing our consideration of ethics and IT with a "theoretically strong perspective" and reflects the historical expansion of the world of humans to include information: "the time has come to translate environmental ethics into terms of infosphere and information objects" (Floridi, 2002: 291).

As a result, Floridi presents two key theses concerning Information Ethics. The first is that information objects are moral agents - that is, they act in ways which can be considered moral or immoral, rather than just in an instrumental manner.

Humans (as information objects) fit this claim in a commonsense manner; but Floridi, of course, has defined many other things as information objects. This includes "the new artefacts, made possible by ICTs" which "may now be endowed with capacities to plan, initiate, hinder, facilitate or constrain actions and behaviours that often have significant moral value and profound consequences (what has been called ''artificial intentionality'') for our and future generations and the whole environment, thus becoming increasingly less distinguishable from moral agents" (Floridi and Savulescu, 2006: 155).

Thus he is able to state that a moral agent is ''any entity...capable of producing information phenomena that can affect the infosphere''. (Floridi 1999: 44; see also Floridi and Sanders 2001).

The second key point for Floridi is that information objects are not just moral agents but have "intrinsic moral value" and thus are the objects of moral action (what he terms "moral patients") (Floridi, 2002: 290). That is, just as we must show respect and care to humans as part of ethical conduct, Floridi argues we owe the same obligation to all information objects, including those that might be considered, from a human perspective, to be information "only" or otherwise artificial.

From these two contentions spring many and varied consequences, if we accept them. To illustrate the impact of this radically different kind of ethical approach, let me provide an example of each: In the first case, we might see how software itself (rather than its creators) acts as a moral agent. In other words, a virus that propagates over the Internet and does harm to the users of that network is itself an unethical actor, breaching moral conventions because of its failure to treat users (and indeed their computers) with moral respect.

Conventionally, we might ascribe the moral failing to the designer or perpetrator of the virus: Floridi's work focuses on the software itself.

Perhaps more significantly, though, Floridi is particularly interested in the way in which all software might function in the ethical domain - not just extreme examples of harmful or illegal software.

Thus, Floridi would ponder the ethical dimensions of the now near-ubiquitous utilisation of PowerPoint presentations which, because of the way that application functions, tend to promote a particular kind of thinking and communication.

Perhaps for that reason and suggesting the power of Floridi's insights, noted design thinker Edward Tufte (2003) has described PowerPoint as 'evil", and moved Don Watson (2004), former Keating speech writer, to criticise it in similar terms.

But Floridi is not just interested in the moral actions of software (as one type of information object, along with many others) in shaping the world. His approach also makes us think carefully about what we do to information.

When we characterise all information - all objects in the infosphere - as subject to ethical consideration, we might thereby gain guidance on what to do, for example, when thinking about deleting those e-mails which could later, have unpleasant consequences should they be found.

Floridi's position is simple. In the words of Nuyen: "What is at issue is the integrity and well-being of information entities in the infosphere.... Justice and fair dealing will preserve and promote their integrity and well-being. Since integrity and well-being vary inversely with the entropy of information, any activity that increases entropy amounts to unjust and unfair dealing [and]... entropy can take one or more of many different forms, such as inconsistency, instability, loss or destruction, unreliability, inaccessibility, and so on" (2004: 190).

Therefore, unethical conduct towards information might be understood in terms of the damage that is done to the infosphere. In this respect, then, the decision as to whether or not to delete e-mails must be made on the basis of the ethical value of those e-mails, not on the basis of the consequences of keeping or deleting them.

This approach, especially when stripped of the more abstract philosophical reasoning with which it is presented by Floridi and to which I do poor justice here, is challenging. It takes very seriously the changing condition of being human in highly technologised, informatic world and proposes that humans radically reorient their view of themselves, so that they are no longer the "centre" of that world but appreciate and recognise that who we are and what we do is bound inextricably to the information that we generate and which gives us identity and the machine processes (hardware and software) of information technology.

It has a certain appeal. As Siponen notes: "To my understanding, at the level of the interface and of a certain level of abstraction - albeit computer software are programmed by humans to ''behave'' in a certain manner - software from an ordinary users point of view seems to posses an autonomy" (2004: 284).

"Equally, I would add, we can see value in Floridi's emphasis on thinking about the moral rights of information itself when trying to think carefully about questions of access, preservation, control, reliability and so on.

Floridi's third law of Information Ethics, "information welfare ought to be promoted by extending (information quantity), improving (information quality) and enriching (information variety) the infosphere" has significant superficial attraction (1999:47; see also Nuyen, 2004: 190-191).

However, there are also fundamental weaknesses. Himma, analysing the logic of Floridi's argument, notes, that there is a problem when Floridi states that humans are informational objects while not considering they are also biological ones: in other words, the infosphere is not the totality of our world but, perhaps, one aspect of overlapping spheres, including the biosphere. Himma also notes that it is not necessary for informational objects to be moral agents to promote the moral respect towards such objects (2004: 145-159).

Siponen also raises some pragmatic objections. He states: "If we go along with [this approach]...it would mean that we can, once again, start blaming computers for our mistakes. In other words, we can claim that 'I didn't do it - it was a computer error' " (2004: 286). Siponen identifies that there is, in fact, a deeper cause - the intentional, reckless, negligent or entirely innocent manner in which the programming was written.

Equally, I would add, granting moral autonomy to software and information systems provides also humans with alibis for their own inadequacies and failed moral choices in using that software.

I am not persuaded that Floridi's approach provides us with the secure foundation for the very real challenges that we face in understanding, accommodating and utilising information technologies in the best possible manner. Floridi considers human society to be strongly determined by the technologies that it produces (or, more accurately, that produce our society).

As a result he is pessimistic about the control we might exert over technology without a new kind of ethics, as evidenced by his comments about hacking (Floridi, 1999). He proposes to accept technological determinism and presume that the development of information technology thus determines correct basis for ethical thought.

Since technological determinism is a product of the manner in which humans themselves construct and understand that technology as a phenomenon, we can largely understand 'technological" determinism as an ideological formation that represents particular human social interests, rather than the ultimate effects of technology itself.

Equally, as Himma has noted, humans are not just informational objects in that certain qualities and characteristics of what it is to be human cannot be readily reduced to a set of information attributes without doing damage to the variety and diversity of the human condition.

These concerns are amplified when we think about how to actually undertake the vital step of moving from abstract reflection to practical action in the promotion of ethical conduct. There is real danger in understanding humanity as if it were somehow equivalent to information processing machinery. Just as many decades were spent fruitlessly attempting to create "artificial intelligence" by presuming computers might think like humans, so now we should be just as wary of assuming people can become intelligently artificial.

As Takenouchi puts it: "If we proceed with this idea alone, the conclusion will be techno-centrist: 'The better a person can process or search for information, the higher that person is evaluated as an excellent machine. At the same time, the weaker one is in processing or searching information, the lower one is evaluated as a machine.'" (2006:189)

That said, Floridi's approach provides two very important ideas which should guide our thinking about ethics and information technology: First, Floridi's turn away from humanity, towards an info-centric view of ethics, reminds us that, in very real terms, there is no longer an easy separation between "us" (humans) and "them" (machines).

This point is evidenced, simply, by the degree to which people that regularly use computers, mobile phones and the like, and experience a sense of loss without them: they have lost some part of themselves. We can also see it in the way humans now exist as "data entities" in numerous, interlinked databases.

Thus, acting ethically in relation to information, and information technology, should not give up on the anthropocentric foundation of moral choices, but should understand that "being human" now involves a large amount of also "being technology".

Equally, while I would argue that Floridi too easily assumes that technologies "control" the development of society, it's also clear that we often too easily presume that humans are always "in control" of the technology.

The comedy program Little Britain humorously, but very accurately, represents this duality. In one skit on that show, an unhelpful travel agent deals with her bewildered customers, by clattering the keyboard of her computer as she seeks information and responds to all requests with a deadpan... "Computer says no...".

Is it the computer, we wonder, or she that is in control?

The ethical challenge, then, is not to presume either computers or humans are in control of the other, but to recognise that the exercise of arbitrary power - the real problem which ethics seeks to address - is now encoded within computing.

And, at the moment of either empowerment or disempowerment, human behaviours and outcomes are very often the consequence of pre-set technological processes, rather than their cause.

Dr Matthew Allen is an Associate Professor at Curtin University of Technology, Perth. He founded the Internet Studies program there in 1999 and is currently researching broadband use in Australia and the development of Web 2.0 as a creative cultural force.

He is currently President of the Association of Internet Researchers. This article elaborates on some issues that Matthew presented recently on a national lecture tour for the Australian Computer Society.

References

Carbo and Almagno. (2001). Information Ethics: the duty, privilege and challenge of educating information professionals. Library Trends, 49.3: 510-518.

Floridi. (1999). Information Ethics: On the philosophical foundations of computer ethics. Ethics and Information Technology, 1.1: 37-56.

Floridi. (2002). On the Intrinsic Value of Information Objects and the Infosphere. Ethics and Information Technology, 4.4: 287-304.

Floridi and Saluvescu. (2006). Information ethics: Agents, artefacts and new cultural perspectives. Ethics and Information Technology, 8: 155-156.

Floridi and Sanders. (2001). Artifical Evil and the Foundation of Computer Ethics. Ethics and Information Technology, 3.1: 55-66.

Himma. (2004). There's something about Mary: The moral value of things qua information objects. Ethics and Information Technology 6: 145-159.

Macavinta. (1999). RealNetworks Faced with Second Privacy Suit. CNET News.com. http://news.com.com/RealNetworks+faced+with+second+privacy+suit/2100-1001_3-232766.html (accessed 11 September 2007).

Noble and Marchetti. (2005). Criminal Records in Victoria: proposals for reform. Melbourne: Fitzroy Legal Service. http://www.job-watch.org.au/jwpublications/crvpr0706.pdf (accessed 11 September 2007)

Nuyen. (2004). Lyotard's postmodern ethics and information technology. Ethics and Information Technology 6: 185-191.

Singer. (1975). Animal Liberation: A new ethics for our treatment of animals, New York: Random House.

Siponen. (2004). A pragmatic evaluation of the theory of information ethics. Ethics and Information Technology 6: 279-290.

Takenouchi. (2006). Information ethics as information ecology: Connecting Frankl's thought and fundamental informatics. Ethics and Information Technology 8: 187-193.

Tufte, Edward. (2003). Powerpoint is Evil. Wired, 11.9, http://www.wired.com/wired/archive/11.09/ppt2.html (accessed 11 September 2007).

Walstrom. (2006). Social and Legal Impacts on Information Ethics Decision making. The Journal of Computer Information Systems, 47.2: 1-8.

Watson, Don. (2004). Death Sentence, The Decay of Public Language. Milsons Point, N.S.W. : Knopf.