05Jun The Tablet: Seán McDonagh – Virtual Friends

Seán McDonagh writes in The TabletVirtual Friends – Artificial intelligence and automation are transforming the way we live and work. Responding to the social, economic and religious impact these new technologies are making will be the defining challenge for the Church over the course of the next few decades.

Link:

https://www.thetablet.co.uk/features/2/20109/robot-ethics-and-virtual-friends

 

Full article:

A ROBOT developed in Japan called “Chapit”, which looks like a mouse, can sit at the bedside of a nursing home resident and engage them in rudimentary conversation. Another robot, “Robear”, has the appearance of a bear, and can lift a patient from the bed to a wheelchair and, later, return the patient to bed. “Palro”, a small humanoid robot, can run quizzes or lead a group of elderly people in exercise routines. At the Shintomi nursing home in Tokyo a robot called “Pepper” is used to care for and entertain the residents; it can also monitor the corridors during the night, so that employers do not have to take on security staff. The creators of these machines hope that they will be used in nursing homes across the world.

In November 2017 researchers at Trinity College Dublin launched a 5ft-tall robot called “Stevie”, which is able to engage in simple conversation and arrange video calls with family members. Its inventors believe it can reduce boredom and stimulate mental activity in older people. It can also remind them when to take their medication, which is particularly important for people who have memory issues. Stevie can also be programmed to recognise what is normal and abnormal behaviour for an elderly person and, if something goes wrong, it is able to notify a carer or relatives.

In 2019 Stevie II was developed by the same robotic and innovation lab at Trinity College and unveiled as Ireland’s first socially assistive robot. It is designed for use in care homes, and can also help people with disabilities who live alone. Stevie II uses advanced sensing technologies that include laser rangefinders, depth cameras and tactile sensors to help it perform various tasks.

PROFESSOR Conor McGinn, one of the scientists who designed the robot, is aware of the ethical and psychological issues involved where, for example, elderly residents in nursing homes might tend to form a deeper attachment to their robot than to their family or friends. McGinn sees other ethical issues surrounding this new robotic technology. He believes that if you build realistic, human- like robots you are basically deceiving people.

Alan Winfield, professor of robot ethics at the University of the West of England, Bristol, has drawn up a list of ethical principles for engineers and programmers who are designing robots that might be used in nursing homes. Winfield serves in a global initiative that is attempting to set ethical principles to be applied to all AI and autonomous systems, such as driverless cars, drones, medical diagnosis systems and search engines. These principles maintain that such systems should not infringe on a person’s human rights and that their functioning should be transparent. Winfield believes that robotics and AI have brought many benefits to humans but engineers and programmers must create robots that benefit humankind and the Earth, rather than merely satisfy their scientific curiosity or their desire to make money.

Winfield would agree with McGinn that a robot should never be designed to deceive: their mechanical nature should be obvious and transparent to their users, and it must always be clear that even though robots are highly sophisticated machines they do not possess human qualities such as understanding, concern and empathy.

The last Japanese census in 2015 showed that one third of the country’s homes are occupied by only one person. Companies have responded by creating “waifu” robots based on popular anime characters, squarely aimed at lonely young salarymen. Their first task is to wake their owner up gently every morning. Then, they remind them to bring their laptop to work. The “waifu” listens to the weather forecast and, if it is raining or threatening to rain, the owner is encouraged to take an umbrella. While the owner is at work, the robot sends affirmative messages and makes sure that the heating is turned on in the house before they arrive home in the evening. When they do, the robot greets them and tells them what is worth watching on TV. Finally, the robot wishes them good night.

Most of these companion robots are presented as female, and the interaction mirrors the misogynist bias in many cultures. Further- more, a dependent relationship with a robot may make it difficult for a person to have trusting relationships with other people.

ADVANCES IN the clinical use of artificial intelligence will also have a profound effect on medicine and on the global medical work- force. To take one example, machine learning has proved to be very effective in scanning diagnostic images in the search for cancerous growths. According to the World Health Organization (WHO), 400 million people do not have access to the most basic medical services. WHO hopes that AI will make affordable healthcare accessible to all people, not just those who live in wealthy countries.

Since 2018, when a person calls the emergency services in Denmark, two things happen in tandem. A human voice answers the phone, but an artificial intelligence assistant also listens to the conversation. This device listens for signs of a potential heart attack as soon as the patient begins talking. It employs algorithms called neural networks, which consist of many different computation layers that act in a somewhat similar way to neurons which are connected in the human brain. While the patient who made the call is answering questions about their age, physical and medical condition, the computer is eavesdropping, and uses speech-recognition software to examine the patient’s words and tone of voice. Then, the computer issues an observation about the likelihood of a heart attack.

In medical trials, the software detected cardiac arrests accurately in 93 per cent of cases. According to a study carried out by the University of Copenhagen, a human health official at the scene will diagnose cardiac arrest failure in just 73 per cent of cases. Furthermore, the computer reaches its conclusions in 48 seconds, which is half a minute faster than a human diagnosis. This device has the potential to reduce fatalities from heart attacks and strokes, major causes of death in many developed countries.

AI could also have a profound effect on the work of doctors, nurses and surgeons. Some robots are currently performing eye operations that are too delicate for a human surgeon to perform manually. During 2018, in Eindhoven in the Netherlands, 12 participants in a trial needed to have a membrane removed from their retina, and this involved cutting several cells which were inhibiting their vision. The delicate operation was performed on six of the patients by highly trained eye surgeons; the other six had the operation performed by a robot. While all the operations were successful, it was found that patients who had had the robotic surgery experienced less damage to blood vessels, especially at the back of the eye. The reason for this is that even the best surgeons do not have perfectly steady hands, which can lead to slight vibrations at the tip of their instruments. Though this is tiny, there are parts of the retina, such as the inner limiting membrane, which are only 0.02mm thick. It would be very difficult for a surgeon to apply the tip of a needle to these parts without causing more serious damage. On the other hand, a surgical robot operates at 0.01mm, making them very useful for undertaking these types of operations.

AT THE JUNE 2018 Digital Health London Summit, Ali Parsa, founder of the digital healthcare company, Babylon, argued that mobile technology, coupled with AI, could make universal healthcare available to people globally, partly by replacing doctors with AI systems, saving costs. He claimed that hospital overcrowding will worsen if we rely on humans alone to provide medical services, and the use of AI and modern technologies will be essential if everyone is to receive good and competent medical treatment. But Parsa believes that doctors and other healthcare professionals will still have a central role, even in the era of these new technologies.

The one thing we can be sure of is that this technology will bring more and more AI into medicine over the next decade or two. Some commentators are now speculating that these new technologies will reduce the number of health and care workers in the future.

AI, robots, drones and 3D printing are undermining the reality of steady employment for everyone. There is mounting evidence that current trends in employment are eroding job satisfaction. When work becomes more automated and inflexible, employees have less control over their workload. Pope Francis is clear that replacing humans with machines is not a good thing. In Laudato Si’ he writes: “The goal should not be that technological progress increasingly replaces human work, for this would be detrimental to humanity.”

Preparing for and responding to the social impacts of AI and automation will be the defining challenge of the next decade or two. It would be disastrous if it takes the Church 30 or 40 years to respond to the impact these new technologies are making. As we enter this fourth Industrial Revolution, we have a short window of opportunity to ensure that these new technologies are aligned with the open, inclusive and dynamic world that we want for all people and for all Creation.

Seán McDonagh is a Colomban missionary. He has written numerous books on ecology and theology, including To Care for the Earth (1986). His latest book is Robots, Ethics and the Future of Jobs (Messenger Publications, £18.95; Tablet price, £17.05).

 

Leave a Reply

Keep the following in mind when writing a comment

  • Your comment must include your full name, and email. (email will not be published). You may be contacted by email, and it is possible you might be requested to supply your postal address to verify your identity.
  • Be respectful. Do not attack the writer. Take on the idea, not the messenger. Comments containing vulgarities, personalised insults, slanders or accusations shall be deleted.
  • Keep to the point. Deliberate digressions don't aid the discussion.
  • Including multiple links or coding in your comment will increase the chances of it being automatically marked as spam.
  • Posts that are merely links to other sites or lengthy quotes may not be published.
  • Brevity. Like homilies keep you comments as short as possible; continued repetitions of a point over various threads will not be published.
  • The decision to publish or not publish a comment is made by the site editor. It will not be possible to reply individually to those whose comments are not published.

 


Scroll Up