Skip to main content

When AI argues back

An audience watches as Dr Ovadia debates with AI

writer icon Hina Javed     IBM   |   Tech     🕐 15. Aug. 2018

In June earlier this year, Noa Ovadia was engaged in a heated debate at IBM’s Watson West site in San Francisco. As Israel’s debate champion in 2016, she is no stranger to an informed argument. This debate was weighing out the pros and cons of subsidising space exploration.

Despite Ovadia’s credentials and accolades, the debater at the other end presented its views without breaking a sweat. The opponent’s lack of nerves and robotic precision were with good reason. It was not human, it was IBM’s Project Debater.

“Just think about that for a moment,”
wrote IBM Research Director Arvind Krishna in his blog which chronicled the event. “An AI system engaged with an expert human debater, listened to her argument, and responded convincingly with its own, unscripted reasoning.”

Interactive Artificial Intelligence (AI) is new, but by no means is it an unprecedented phenomenon.

One of the poster-children for interactive AI is BINA48, who over the years has developed a following thanks to widespread media coverage. The humanoid robot has also appeared on television shows such as the awarding-winning docuseries The Story of God.

Martine Rothblatt paid $125 000 to create the robot as a lifelike version of her living spouse Bina Rothblatt.

The robot is equipped with 32 motors to help it communicate with humanlike expression. The robot’s skin is made of Frubber and the motors help her frown, smile and even look confused, depending on the nature of the conversation.

BINA48 can answer questions, has goals, and says in the first episode of The Story of God that it “hopes to be human one day” so it can go for a walk in its garden.

Too close for comfort? 

The field of interactive artificial intelligence has progressed leaps and bounds since BINA48’s invention. Robots are now more developed in terms of their intelligence and this has led people to address two important, yet unaddressed issues. 

First, is the risk of humans becoming too dependent or emotionally attached to machines. The second is the implication of confusing artificial and real intelligence.

“Caregiving, romantic, and peer or teammate human-AI/robot roles will probably lead naturally into some level of human attachment,”
says Dr Julie Carpenter, who has published a book on the relationship between robots and humans.

Therapeutic use

In her conversation with Forbes magazine, Dr Carpenter says that she is a roboticist. She also had a PhD in Learning Sciences from the University of Washington.  

She explains that in some cases, the therapeutic use of a robot for companionship could help improve the lives of people with certain conditions. On the other hand, she warns that the emotional attachment is a “double-edged sword”.

“For all the pleasure emotional attachment to something can bring, other outcomes can be loss, or loneliness,”
Dr Carpenter cautions. 

She says that an attachment to a machine, especially one that can interact, converse and even seem to care, can breed the desire in humans to give back. “It makes people less likely to want to lose that thing or be separated from it.”

The roboticist outlines that if the machine is damaged or lost altogether, it could have an emotional impact. The extent of the trauma depends on the individual.

Making communication easier 

On the other hand, Silicon Valley Robotics Managing Director Andra Keay sees AI robots as a great way for humans to avoid interaction with other people when they are not comfortable. In her talk at the Web Summit Conference at Dublin, she used the example of a particular robot named Mabu.  

Mabu interacts with patients and passes on data to doctors, thus eliminating the need for human-to-human engagement. This, she believes, can prove particularly effective if people are uncomfortable discussing certain medical conditions face-to-face.

“Anything with a screen, in fact, anything with speakers and connectivity becomes a gateway for many other people and can offer multiple levels of interaction. Welcome to life with robots,”
she said at the conference.

Widespread use of AI
Individual interaction with robots is one facet of artificial intelligence. Tech firms continue to make giant strides in bringing the cognitive abilities of their creations on par with humans.

There are general fears that robots, armed with the ability to think, will start replacing humans. However, Matthew Lieberman, who is on Forbes’ Technical Council, thinks otherwise.

“AI encourages a gradual evolution in the job market which, with the right preparation, will be positive,”
he asserts. “People will still work, but they’ll work better with the help of AI. The unparalleled combination of human and machine will become the new normal in the workforce of the future."

Where to draw the line?
However, a 100-page report, written by 26 artificial intelligence experts from 14 institutions including Cambridge, Oxford and Stanford universities, warns that advances in the field need to be checked. The report focuses on AI attacks humans are likely to face if they do not put up adequate defences.

While they do not warn of nuclear warheads launching themselves in order to end civilisation as we know it, the experts believe the cost of simple criminal acts, such as hacking, will be lowered by the scalable use of AI.

The report cautions that “ordinarily [such tasks] require human labour, intelligence and expertise. A natural effect would be to expand the set of actors who can carry out particular attacks, the rate at which they can carry out these attacks, and the set of potential targets.”

A lack of research

It also warns that any malicious intent or attacks, aided by artificial intelligence, can be especially effective, finely targeted, difficult to trace and more likely to exploit vulnerabilities in the systems. The experts conclude that one of the most alarming factors is this the fact that there is little research, or regulation, to combat rogue weapons using artificial intelligence.

AI has obvious advantages. It is for humans, however, to keep a check before they get either too emotionally attached to a single caregiving robot, or before they put too much faith in the widespread use of artificial intelligence.

We believe that information should be free and will therefore never put up a paywall.

If you like reading our reports about the Scandinavian business scene and would like to donate towards the upkeep of the site, we would be very grateful. Click here to donate.