Many issues surrounding robotics, artificial intelligence

Source: Credit Suisse. From left: Professor Thalmann, Graylin, and Dr Hechler-Fay'dherbe with Greg Williams, Editor, WIRED (right).

Robots are nowhere near taking over the world, let alone interacting on a daily basis with humans. Artificial intelligence (AI) is progressing by leaps and bounds, and will soon affect our lives intimately. These were the opposing views at a panel called Technology at the service of humans during the 2018 Credit Suisse Global Megatrends event.

Professor Nadia Thalmann, Director, Institute for Media Innovation (IMI), Nanyang Technological University, Singapore and Director, MIRALab, University of Geneva, Switzerland said robots still lack awareness of their environment. "(A) robot is unable to be aware of where the robot is and all the people who are here, so we still have a long way to go," she said.

"There is still a lot of research to be done for a robot to be aware of its environment," she said. "In five years, there will incrementally be companions who are aware how you feel, how you are and react in an appropriate and personal way."

Alvin Wang Graylin, China Regional President of Vive, HTC disagreed. Vive is HTC's virtual reality (VR) business unit. He believes that it could be a matter of hours to go from a system with average human intelligence to "super AI".

"It could be 100 to 1,000 times more intelligent, and can solve problems we can't solve right now," Graylin said of a super AI, noting that the AlphaGo Zero AI had taught itself without learning from any masters and was able to beat the original AlphaGo by a large margin.

Graylin said, "There is a lot of fear that AI robots are going to take my job. For some categories of jobs it's going to happen sooner than others."

People could benefit if certain jobs are replaced, he said. "If it's repetitive work people could probably put their minds to useful labour."

"Get people to be familiar (with AI) and accept and adapt. That's really the first step," he advised.

Dr Nannette Hechler-Fayd’herbe, Global Head of Investment Strategy and Research, International Wealth Management, Credit Suisse, made the distinction between AI that are working on specific situations and AI which can make predictions and are able to react to unforeseen patterns. "We are really a long way away, maybe 15 years, maybe less than that," she said of the latter category of AI.

Professor Thalmann said social robots will have the biggest impact on our lives, and suggested the audience invest in robot manufacturers. "(A social robot) will be our companion in the future, not only at home," she said. "It can understand (you), be personalised to (you) and really react and do a lot of things."

The conversation also touched on VR, which could be how we interact with AI in the future. Graylin believes that the way the AI Jarvis interacts with his owner Tony Stark in the movie Iron Man will be the future. The two speak to each other and Stark uses gestures to manipulate holograms that Jarvis presents. "This new interface will become our new interface to data, to each other, to machines," he said.

Dr Hechler-Fay'dherbe said VR would take root in three phases, with slow adoption when use cases are limited, in this case to gaming; a growth phase when the technology becomes the mainstream way to interact, possibly with holograms, followed by  "a breakthrough to the frontier to become the mass medium".

"This is really when the application case is really starting to become explosive," she said. "We are probably somewhere in the second stage."

While AIs can interact today through wireless devices, it could be some form of glasses in five to 10 years, and later still, a contact lens that is connected to the cloud, Graylin predicted. "VR, if it's done right, will change most of the fundamental things we do today," he said.

He said the movie Ready Player One, which sets the action in a virtual reality (VR), will create social acceptability for VR glasses. "Google Glass started off hot and then became a failure," he said. "There was a social stigma (around) someone wearing a computer on their head and people worried about privacy issues."

"After (people) watched the movie they were twice as likely to buy a VR headset," he shared.

Use cases for VR could include making Nobel prize winners accesible to children anywhere in the world, or taking a virtual vacation with the family. "This is going to make what used to separate the 'haves' and 'have nots' disappear," he said. "This is going to be accessible to anybody who can afford a few-hundred-dollar device."

Dr Hechler-Fay'dherbe warned about addiction. "We anticipate a lot of studies on how these virtual realities going to impact behaviours," she said.

Graylin responded: "The desire to escape from the real world exists anyway. It's about how to get people to manage their desires than an inherent evil in the technology."

On the flip side, Graylin noted that AI could overpower everything else. Professor Thalmann said, "The software can control everything...It is up to us to decide, because in the end we are the ones who write the software."

Dr Hechler-Fay'dherbe also reminded the audience to be mindful of maintaining a diversified portfolio in tech stocks. She listed robotics, AI, health technology, VR and AR as possibilities. "Be clear how fragmented these developments are," she said, pointing out that regulatory compliance may also affect whether technologies are worth investing in.

The session added with a demonstrate of the robot Nadine, which is modelled on Professor Thalmann. The robot could handle questions put to it in different languages and different accents, and blinks and smiles as well.

Other polls held during the event included:

How much worldwide revenues for AI would be by 2025, with the choices being under US$10 billion, around US$20 billion and over US$35 billion. About 67% of respondents voted for the highest value, US$35 billion.

"AI is the most significant technology of the coming decade". Approximately 87% voted yes.

"My organisation has a clear strategy for the implementation of AI". About 78% said no.

"Are governments prepared for the structural shifts that will occur in mature economies in the coming age of automation?". The answer was no, from 81% of the audience.

"Machines will develop intelligence superior to that of human beings by 2040". Roughly 78% said yes.

Hashtag: #CSMegatrends