No sex please, robot, just clean the floor
Ed Habershon and Richard Woods
THE race is on to keep humans one step ahead of robots: an
international team of scientists and academics is to publish a
code of ethics for machines as they become more and more
sophisticated.
Although the nightmare vision of a Terminator world controlled by
machines may seem fanciful, scientists believe the boundaries for
human-robot interaction must be set now before super-intelligent
robots develop beyond our control.
There are two levels of priority, said Gianmarco Verruggio, a
roboticist at the Institute of Intelligent Systems for Automation
in Genoa, northern Italy, and chief architect of the guide, to be
published next month. We have to manage the ethics of the
scientists making the robots and the artificial ethics inside the
robots.
Verruggio and his colleagues have identified key areas that
include: ensuring human control of robots; preventing illegal use;
protecting data acquired by robots; and establishing clear
identification and traceability of the machines.
Scientists must start analysing these kinds of questions and
seeing if laws or regulations are needed to protect the citizen,
said Verruggio. Robots will develop strong intelligence, and in
some ways it will be better than human intelligence.
But it will be alien intelligence; I would prefer to give priority
to humans.
The analysis culminated at a meeting recently held in Genoa by the
European Robotics Research Network (Euron) that examined the
problems likely to arise as robots become smarter, faster, stronger
and ubiquitous.
Security, safety and sex are the big concerns, said Henrik
Christensen, a member of the Euron ethics group. How far should
robots be allowed to influence peoples lives? How can accidents be
avoided? Can deliberate harm be prevented? And what happens if
robots turn out to be sexy? The question is what authority are we
going to delegate to these machines? said Professor Ronald Arkin,
a roboticist at the Georgia Institute of Technology in Atlanta.
Are we, for example, going to give robots the ability to execute
lethal force, or any force, like crowd control? The forthcoming
code is a sign of reality finally catching up with science fiction.
Ethical problems involving machines were predicted in the 1950s by
the science fiction writer Isaac Asimov whose book I, Robot was
recently turned into a Hollywood film. The Terminator and Robocop
series of films also portrayed mechanical law enforcers running
amok.
Present robots perform more mundane tasks: the most common consumer
robots in Britain include self-guided vacuum cleaners such as the
Scooba, lawnmowers such as the Robomow and childrens toys such as
Robosapien.
But far more sophisticated machines are being developed. The
National Health Service has used a robot called da Vinci to perform
surgery at Guys and St Thomas NHS Foundation Trust in London. In
Japan, human-like robots such as Hondas Asimo and Sonys Qrio can
walk on two legs. More advanced versions are expected to be
undertaking everyday domestic tasks and helping to care for the
elderly in as little as 20 years.
I would hope they would always be subordinate, said Brian Aldiss,
the science fiction writer. But one will no doubt come to rely on
them deeply. Aldisss short story Supertoys Last All Summer Long
was the basis for the Steven Spielberg film AI, which addressed the
subject of whether androids that have become as intelligent as
humans should be denied equal rights.
Other dilemmas may arrive sooner than we think, says Christensen.
People are going to be having sex with robots within five years,
he said. So should limits be set on the appearance, for example, of
such robotic sex toys? The greatest danger, however, is likely to
lie with robots that are able to learn from their experiences. As
systems develop, robots are likely to have much more sophisticated
self-learning mechanisms built into them and it may become
impossible to predict exactly how they will behave.
My guess is that well have conscious machines before 2020, said
Ian Pearson, futurologist-in-residence at BT. If we put that in a
robot, its an android. That is an enormous ethical change.
To critics who scoff that intelligent robots are a long way off,
the roboticists easily riposte that machines can already exert
surprising influence over our lives think about the influence of
the internet.
Keeping control
New robo-ethics recommendations
Safety Ensure human control of robot
Security Prevent wrong or illegal use
Privacy Protect data held by robot o Traceability Record robots
activity
Identifiability Give unique ID to each robot
Isaac Asimovs laws of robotics
Robot may not injure human or, through inaction, allow human to
come to harm
Robot must obey human orders, unless they conflict with first
law
Robot must protect itself if this does not conflict with other
laws
