Oslo Business Forum - Articles

Kate Darling: The potential of robots is to help humans do better jobs

Written by Emily Northway | Jan 7, 2019 9:18:52 AM

 

How will the future interaction between humans and robots play out? And what should business leaders be aware of in times with an increasing amount of robots? MIT researcher, Kate Darling, shares her tips on how business leaders can successfully meet a robotic future.

This article will cover these questions:

  • How do humans interact with robots?
  • How are robots changing the way we live, work and lead?
  • What should you as a business leaders keep in mind when your company adapts more technology and continues moving into Industry 4.0?

 

Kate Darling works as a researcher at the MIT media lab where she studies human-robot interaction. More specifically, how people interact with robots and how they psychologically relate to them. 

“I’ve always loved robots. I used to read a lot of science fiction as a kid,” she says.

But instead of becoming a science fiction author or studying technology and robotics, she decided to go to law school where she got a law degree.

“I wound up with a job at MIT that let me work closely with robotics. I realized that there are people that understand the technology really well, and there are people who understand social sciences, law policy, and ethics really well. There are a lot of questions that are at the intersection of those,” Darling says.

Robot technology is moving faster than ever before. From huge factories into new spaces where it’s interacting with people. “It’s kind of a new era. It’s the era of human-robot interaction that we’re about to enter,” she says.

But what is the definition of a robot? There are some semi-official definition, but the one Darling is using the most is:

«A physical machine that can sense the world, that can make some sort of autonomous
decision based on what it sensed, and can act physically on the world.»

“If you go down into the details, it's pretty difficult to define what a robot is. Sophia, Hitch, the No Isolation-robot have all very humanoid shapes. But an arm that makes a car in a factory can also be a robot.”

«Robots are still really dumb»

Even though robots have been present behind the scenes – in factories and production lines for a few decades, robots are now entering more social spaces, according to the MIT researcher.

“Robots are still really dumb. They’re not sophisticated.”

“What’s interesting is that robots are starting to interact with people and becoming more social. Even though robots don't have intelligence anything near what we would compare to human intelligence, we're starting to engage with robots as though they are pets, or social actors because we project onto them more intelligence than they actually have - more capability. It can be fun in some cases and problematic in others,” she adds.

“In what cases is it problematic?”

“There is the concept of automation bias where we trust machines too much. We trust their decision-making, or we trust that an algorithm is neutral and not biased. Oftentimes that's not the case, because algorithms learn from data and our data sets are all biased,” Darling explains.

Darling uses the doll called Cayla as an example, where the Norwegian Consumer Council have been criticizing the doll for breaking Norwegian laws and «fail miserably when it comes to safeguarding basic consumer rights, security, and privacy

“Cayla would record all the conversations that it was having with children – which is problematic enough on its own. But then also the terms of service said that the company can use the information in advertising and give it to third parties. We're starting to reveal information to them.   But the information is not being kept by a friend. It’s being kept for a corporation, that doesn't necessarily have our best interests in mind. So those are just some examples for why trusting robots maybe sometimes is not a great idea.

According to Darling, there are also great cases for having robots that we develop trust with. She uses an example where robots are working with autistic children to engage them in new ways. Robots are also being used therapeutically, for example as animal therapy replacements. The Norwegian telepresence robot, NoIsolation goes into schools and helps children connect with their classmates.

«The potential of robots is to help people»

As robots are entering social spaces, they will have use cases both in our private lives and in businesses. Darling believes that all industries will see an increased amount of robots.

For the last few years there have been debates on robots entering the health industry to replace human care have been going on.

Darling believes that numerous jobs will be taken over by robots, but also wishes to nuance the future a bit.

“I actually think it’s a problem that we’re trying to develop humanoid robots to take over human tasks. That’s not useful. If you’re looking at care settings – for example, taking care of old people – there’s so many ways that human care is much more valuable in this context. Robots just aren't good at a lot of things that humans are good at – and they won't be for a long time.”

“It’s sad to me that we're trying to replace human jobs when, really the potential of robots is to help people do better jobs and be supplemental to human ability. So I think robots can be great to help as tools in care settings, but I don't think they should be taking the role of the caretaker,” she adds.

Digging into the future of AI

Last year, Google gave a stunning demo of a Google Assistant making two actual phone calls – one to the hairdresser and one to a restaurant.

During the event where Google showed off the new capability of the company’s Assistant: It can make appointments and phone calls on your behalf. You can hear the conversation here:

 

 

These bots are designed to sound really human-like. They make little mistakes and pauses and say, eh, ah. What's going to happen is the hair salon is also going to have a bot making the appointments. And so you’re going to have two bots talking to each other – both trying to sound as human as possible,” Darling says.

But how should we act towards these kind of systems when interacting with them? “We do already act towards them as though they are a little bit alive – if they're well designed,” she says. Adding that some parents has already complained, for example, that their kids won’t say «please» and «thank you» to Alexa.

“Amazon actually had to release a new feature that you can turn on, that makes you say «please» and «thank you» to get something from Alexa, because they realized that these interactions that our kids are having with machines are like other social interactions, and that it can bleed over into their behavioral patterns,” she says.

«Stop comparing AI to human intelligence»

One of the things that worries Darling the most within robotics and AI, is that “we’re always comparing artificial intelligence to human intelligence,” she says.

The MIT researcher believes that the people who are going to be able to make use of the true potential of the technology, are the ones who “understands that, thinking outside of the box and trying to find ways to use this technology to supplement what we're doing and be a partner in what we’re trying to achieve, rather than just recreating what we already have. Those are going to be the winners because that's the true promise of the technology.”

“If there's anything that I could say to business leaders is stop comparing AI to human intelligence and think it's out of the box,” she says.

 

 

She uses the baby seal robot called Paro as an example. You might have seen it? It was actually first exhibited in 2001. This robot is used in nursing homes with dementia patients. It appears as cute and fluffy for many, and people love it.

“It also freaks people out because everyone is like: «I can't believe you're giving old people robots instead of human care, that's terrible»,” Darling says.

“But that’s not what this robot does. This robot doesn't replace human care, it's supplements human care. It’s a replacement for animal therapy. So thinking of the ways robots can be used to help caretakers the way that we've used animals in the past. That's really where you want to go,” she adds.

This is what you need to enter a robotic future

When entering a world with increasingly more AI and robotics, Darling believes the most important thing for business leaders is to develop a better understanding of the capabilities of the technology – along with the potential and limitations.

“I think we're very caught up in the science fiction ideas of what artificial intelligence and robots can do, or can't. Therefore, I think people sometimes overestimate or underestimate what the technology can do. Having a better educated understanding would help us,” she says.

Adding that business themselves need to understand the technology. “I think it’s education. It would help if the media were a little bit better at educating people rather than putting a picture of a terminator on every single article about robots,” she says.

Robots are evolving in the age of Industry 4.0. Some industries have already been through substantial technological changes – and there will be more of them.

“I do think that robots are going to disrupt labour markets. I just think it's not as clear as one robot is going to take one person's job. But it's much more complex. I think it depends on the industry and how that disruption is is going to impact people and jobs,” Darling says.

“In some cases it may be a good thing and lead to growth, in other cases in might be more difficult. For example, once trucks are automated, a bunch of truck drivers will be out of work and it might be difficult to relocate truck drivers in a way that it wasn't difficult to relocate journalists when the Internet happened. So, it really depends on the individual industry. I do think there will be more robot and human interaction in most Industries as well,” she added.

 

Kate Darling joined the Future Forecast podcast to discuss human/robot relationships. Tune in!