For most of the general public any talk of artificial intelligence (A.I.) and robotics is more than likely to be dismissed as mere science fiction. It is something for eccentric and geeky scientists and technologists to fantasize about but of no real value or importance to anyone living in the real world where the real challenges of life prevail.
Nevertheless, the advances in A.I. and robotics look set to mean that in a mere 10 years time autonomous systems will probably play a more significant and tangible role in our lives, be it in the form of help around the home or personal care. To be clear, I am not thinking in the first instance of a world similar to that of the one Will Smith found himself apart of in the film, I, Robot, although the obvious implications depicted in the film help to sharpen our thinking in terms of the ethical, legal and social implications for the long term. Instead the prospect of doing life with the assistance of an autonomous system is something that will become more apparent.
A report recently published by the Royal Academy of Engineering (RAEng) on autonomous systems seeks to raise questions and highlight the ethical, social and legal implications (ELSI) such systems present to us. The report offers a range of different types and definitions of autonomous systems. Some systems involve humans maintaining full or partial control whilst others have supervised systems. Automatic systems carry out fixed functions without the need for any human intervention whilst others are truly autonomous and are adaptive, able to learn and make decisions.
Autonomous systems look set to change our lives considerably and revolutionise many aspects of our lives, should we agree to their adoption. Nevertheless, if we took the time to reflect on what autonomous systems represent then we would quickly realise that we have been working with them for some time.
Take for example the lift we use to reach our office in the morning, move between floors in the local department store or simply use to get upstairs in our own homes. Whilst in the early days lifts may have had their own operators today we freely use such systems on our own. Likewise, consider the monorail system which operates without a human driver. Consider the autopilot on the aircraft we flew on to our holiday destination over the summer. Surely that is an autonomous system; a piece of technology designed to help the human pilot in their role and responsibilities. We could go on and consider the common thermostat for controlling the heating in our homes or the anti-lock braking system on our modern cars. The point I wish to make is that there could be a tendency to think that in light of the RAE publishing their report, autonomous systems are a new phenomena presenting seminal sets of questions. In fact they do not. We have lived alongside such systems for much of our lives and for the best part, without even realising that we were.
Sharpening our thinking
Where the value of the RAE’s report can be found is in helping to sharpen our attention and our thinking in terms of where this kind of technology could head given the expanse of its development to date. It is inevitable that as this expansion continues, what were largely unnoticed forms of autonomous systems will lead to far more visible and direct instances of interaction between human and machine or humanoid. It is into this context that the RAE report helps to raise some pertinent and helpful questions.
The RAE report focuses on two main areas of autonomous systems. First, transport and cars that drive, steer and park themselves and secondly personal care and support - "smart homes" that monitor health and well being as well as systems which help to provide companionship for those that find themselves on their own.
Autonomous transport could help pave the way to unlocking the congestion on our roads leading to better traffic management and safer forms of travel. On the other hand, in those cases where some one may be killed by a vehicle which was not manned by a human, who would be responsible for the death? Is our legal system ready to deal with the advent of autonomous systems? The law currently makes a distinction between human operators and technical systems and requires a human to be responsible for an autonomous system. How would the situation change should these technologies enhance or extend human capabilities to compensate for impairment to cognition or motor functioning? Would humans be responsible for their autonomous system or are the two separate? This is similar to the issues encountered with genetic screening and testing. As exemplified in the film Gatacca, insurance companies may not provide cover for those who are seen to be too much of a risk. In the case of autonomous systems, could a trend emerge in the insurance industry whereby insurance is withheld from certain enhanced individuals?
Ageing population
An ageing population is a challenge which faces many countries at this time. Consequently, the prospect of autonomous systems which care for and give companionship to lonely and house bound individuals could ease the burden on healthcare professionals and services. Effective systems could be introduced in order to ensure medication is taken at the right time and the necessary emergency support called for following an accident within the home.
The Japanese, who could be termed world leaders in the field of A.I. and robotics, are already familiar with this kind of situation. It is reported that nearly 22 percent of Japan's population is already aged 65 or older. This has resulted in a plethora of devices, ranging from easy-entry cars and intelligent kettles, being released to respond to this need and fuelling a care-technology market which is reported to be worth in the region of $1.08 billion in recent years[1].
Autonomous systems are proving to be beneficial in seeking to maintain a sense of independence for the elderly by allowing them to stay in the familiar and often cherished surroundings of home instead of having to move into a care home in order to have the care and support around them which they require. The dull and routine household duties could be undertaken by robots as well as helping to keep track of activities which a failing memory could easily forget. Nevertheless, we have to be aware of a possible trade-off. It could become all too easy for the robots once they have proven their worth in these ancillary roles to be left to fulfil the role of constant companion or replacement family member. I can’t really find the time to visit Nan today but she has her robot for companion so she should be OK.
There is something to be said for human and social interaction. The elderly need the untold benefits of love and human contact. This would come from visiting human carers and not robots. One only has to consider the difficulties which remain in trying to ‘teach’ robots to understand humour and spot a joke to realise there is still much to be overcome in terms of creating a robot which is able to replace a human in every sense[2].
Keeping the on/off switch
An automated system which still have a level of human operation involved is somewhat different from an autonomous system which is able to adapt, learn and make ‘decisions’ free from any human intervention. These two examples could be used to serve as two opposing points at either end of a spectrum. What is importance for us to do in the days to come is to determine where we wish to positioned ourselves on that spectrum. Essentially it boils down to whether or not we wish to keep the on/off switch or do away with it completely. Is the idea of an autonomous system which we can still maintain control over by switching it off more favourable to one that we cannot switch off? Does the fact that a human could switch it on and off render the whole idea of an autonomous system useless?
Some argue that autonomous systems are not a major issue and that a laissez-faire approach should be adopted in terms of adopting such systems into our lives. This may spark alarm from many but when one considers the example of the internet and the mobile phone, both examples provide a compelling case of how innovation that may initial go against the grain of the zeitgeist and pioneer something new in terms of new forms of understanding what justice, morality and ethics represent can actually bring many benefits. The critical question is whether or not autonomous systems can be afforded such freedom or whether the level of transformation/disruption is such that they necessitate a greater level of engagement.
In many respects, the issue of autonomous systems is no different to any other major challenge to society which requires us to reflect and think upon the ethics involved. Assuming an Aristotelian position in relation to ethics, Andorno notes that ethics reflects the effort of our reason in discovering whether something is right or wrong, with the aim of promoting the means by which we reach the fulfilment of our tendencies towards the good[3]. In contrast the basic purpose of law is just to ensure that human relationships are governed by the principle of justice, or in other words, that the rights of each individual, as well as the common interests of society as a whole, are guaranteed[4]. Thus we really do need to sharpen our thinking and ethical question of these matters in order to understand what legal instruments are necessary and appropriate for the future.
In this vein, the initial questions posed by the RAE report may seem a little belated and redundant. We appear to be quite happy with autonomous systems being part of our lives. However, as current developments take place and we look to the future we need to review the ELSI involved with future advances in this area, specifically regarding humanoid interaction. It is not so much dreaming up an ethical code for robots than perhaps developing a robot ethics for us that can be adopted in order to begin to comprehend and grasp our future interactions with the next phase of autonomous systems and robotics.
The recent RAEng report 'Autonomous Systems: Social, Legal & Ethical Issues' can be read and downloaded here .
-----------------------------
[1] http://www.foxnews.com/story/0,2933,299677,00.html
[2] http://www.newscientist.com/article/mg19526156.400
[3] Andorno, A. (2009), Human dignity and human rights as a basis for global bioethics
[4] Ibid.
