The promise and peril of AI
You have probably chatted with Amazon’s Alexa or Apple’s Siri. Autonomous vehicles are proliferating in the air, on the water and on land. And you may be familiar with the personalised recommendations of Netflix. But did you know that your bank’s decision to approve or deny your loan application may rely on the judgment of predictive algorithms?
Like it or not, AI already affects, and will continue to affect, our lives. This process of incremental transformation will force us to reassess our most basic moral, social, and legal values.
Imagine the struggle of a judge whose instinct tells him to deny bail to a defendant, yet an arguably more intelligent AI-enabled decisionmaking system suggests otherwise? Would you put your elderly parents in the care of an elder-care bot? Would you want your doctor to use predictive analytics to decide the best treatment for you, or be operated on by a medical bot?
Are you comfortable with the thought that, using pattern recognition and/or autonomous surveillance, Google, Facebook, and others may know more about you than your closest friends and family, and are likely able to anticipate your next move before you do?
AI technologies also hold the promise of enhancing future societies in a number of ways: self-driving vehicles are expected to reduce the number of accidents, smart transportation systems can solve urban density problems, remote communities without access to doctors could enjoy medical services provided by a robot, autonomous surveillance is able to filter out deviant individuals by establishing connections beyond the limits of what’s humanly possible.
Yet, the extent of benefits strongly depends on developers’ incentives, and in any case, the use of AI technologies presents serious challenges. For one thing, competition among firms and countries for technological superiority – a phenomenon often referred to as the AI race – can inflict severe damage by encouraging corner-cutting on governance or safety issues, and may also spark international conflicts.
Programming a self-driving car to sacrifice a sick person in order to save a healthy individual in an unavoidably lethal crash scenario reflects a fundamental moral decision to value the healthy over the sick, as in ancient Sparta. Policymakers and the legal profession will soon have to answer tricky questions like: who is responsible if a patient dies after misdiagnosis by an algorithm; who owns intellectual property rights over various instantiations of AI; who should rightfully own/ have access to data acquired by intelligent transport and autonomous surveillance systems?
Before long, we will all need to use, collaborate with, or rely on the judgment of different AI applications in our professional and private lives. According to widespread perception, we may even find ourselves competing with AI for jobs. While this fear around employment is certainly understandable and to some extent justified, it is important to see that this process will also create many new jobs and spur economic growth in previously unimaginable ways.
AI has tremendous potential to change our lives for better or worse. Ensuring that these powerful technologies are beneficial for humanity will require joint effort and sincere co-operation among all nations and segments of society.
Policymakers must step up and design responsible and sustainable AI policies. To do so, they will not only depend on academic and industry stakeholders invested in AI research and development, but also on the moral judgment and views of ordinary people, who should have a say in these life-changing decisions.
We should not be intimidated by the rapid pace of technological innovation, but rather allow ourselves time to develop policy responses and new solutions in a process of careful deliberation and thorough consultation to adequately inform the general public and make sure their opinions also flow into new policies on AI.
Luckily, the New Zealand AI community is very much aware of these exigencies, as illustrated by AI-DAY, New Zealand’s premiere AI event, held in Auckland a month ago. This successful AI conference provided a stocktake on existing AI technologies and connected and inspired debate between those involved or interested in AI research and development.
Alongside a variety of companies sharing their innovative ideas, and two panel discussions featuring academics and industry representatives, Broadcasting, Communications and Digital Media Minister Claire Curran outlined the Government’s work to prepare New Zealand for the AI era.
Similar positive developments are taking place in many other countries, offering a glimpse of hope that the international community has learned from past mistakes and will at least make an effort to address a truly global phenomenon with coordinated international action.
Whatever happens, we must bear in mind that here, at the dawn of the AI age, we hold both the power and the responsibility to shape AI technologies in humanity’s best interest, and for the sake of future generations, it is imperative that we make the most of this opportunity.
Dr Erdelyi recently took part in a panel discussion in New Zealand’s first AI Forum, AI-DAY.