Dr Hanson said that AI needs to be raised in a “human family”
In the 1982 science fiction cult film Blade Runner, Tyrell Corporation, the organisation responsible for manufacturing androids (or “replicants” as they are called in the film), boasted its androids were “more human than human”. This implied that they not only physically resembled and moved like human beings but also expressed and felt human emotions.
Fast forward about 30 years, rapid advancements in artificial intelligence (AI) research have led to the founding of many companies who could, one day, actually build real-life androids; capable of expressing empathy, joy and, perhaps, even love — just like in the film.
One of these companies is Hanson Robotics, a Hong Kong-based robotics startup led by CEO and Founder Dr David Hanson.
At the Skolkovo Open Innovations Forum, a Russian government-backed technology and innovations conference held in Moscow three weeks ago, Dr Hanson graced the main podium to talk about his creations and evangelise about robotics’ future. As he spoke, one could see the parallels between his beliefs and that of Tyrell Corporation (albeit, far less nefarious).
Sophia the robot
Dr Hanson wasn’t the only “speaker” on the stage; standing with him was Sophia the robot. If that name rings familiar to you it is because Sophia has been making rounds across the world, on TV talk shows such as Jimmy Fallon, and even becoming an honorary citizen of Saudi Arabia.
On her tours and appearances, she typically takes questions (most of them prepared beforehand) from hosts or reporters and gives intellectual or clever responses (sometimes, snide). Sophia can also read face and hand gestures.
Hanson Robotics is not the sole creator of Sophia; she is the handiwork of some of the most advanced developers in AI today. Alphabet Inc, the parent company of Google, is behind Sophia’s voice recognition system. The engine that powers her brains and allows to respond comes from SingularityNET, a blockchain-powered AI marketplace, which is dencentralised marketplace for AI programmes.
Also Read: This is how neural networks are enabling smarter AI products
And it wasn’t just Sophia’s brains that astounded the audience; Sophia possesses life-like skin, which is a “bio-inspired material” called Frubber, explained Dr Hanson. The patented material “is an extremely soft material similar to human flesh,” he said, which enables for natural facial expressions.
Not convinced? Just take a look at this video by Tech Insider — it is eerily convincing.
“We need to treat them [robots] as friends…use data and machine learning to train AI to learn about us…raise AI in a human family and teach them what it means to be human,” he said.
Humans are social creatures, it is not enough that they interact with faceless assistants like Google Home or Amazon Alexa.
According to conventional scientific belief, at least 55 per cent of all communication is non-verbal — humans. Thus, the development of realistic physiology and facial animations that can mimic lifelike facial expressions, can provide for another — and deeper — level of connection.
Dr Hanson broke down the ideal AI into four characteristics: adaptive, conscious, caring, and ethical. The more a robot interacts with people, and the more data they will collect, the more human they will become, ideally.
“Sophia goes out to meet people, they also write messages to Sophia,” said Dr Hanson.”Having so many people engage with Sophia means she learns from many people”, allowing her to draw from a rich pool of varied human experiences.
Upgrades for Sophia already in the works and they will be implemented in different stages. First, Sophia will receive a “grasping arm” that will be “very sensitive”. Next, her computer vision will be upgraded to read facial and hand gestures. By 2019, the goal is for Sophia to walk.
A Chinese version of Sophia, aptly called Ai (which is the Mandarin term for ‘love’), is also currently in development.
Other creations and the future
Sophia may be Hanson Robotics’s main attraction, but there are other robots in its stable. One of them is Professor Einstein, an educational robot that helps kids learn about science. Kids can also tap into Dr Roboto, an open source AI platform, to program new movements or expressions on the robot.
Besides that, Hanson Robotics is also working to develop new robots in partnership with Disney. Hanson Robotics is part of Disney’s media and entertainment innovation programme Disney Accelerator,
But while these robots may serve different purposes, they are all engineered towards Hanson Robotics’ founding goal, which is to infuse the essence of bio-engineering with a complex system of AI.
Also Read: Asia is building the future for better quality of life with smart cities: MIT report
Dr Hanson predicts that within five to 10 years, robot intelligence will surpass human intelligence. Hanson Robotics is ramping up development of its cloud-based AI engine MindCloud, lead by Dr Ben Goertzel, who is also the Founder and CEO of SingularityNET.
The goal is to build an ambitious AI framework, said Dr Hanson. Robots of the future will be capable of “symbolic reasoning”, “whole organism cognition” and “simulated physiology” that will allow for realistic breathing.
“How much consciousness can you fit into the chips — that is the question,” said Dr Hanson.
Then comes the inevitable existential questions, which many theorists, tech experts and founders, such as Elon Musk, are already starting to ask: What will we do when robots become smarter than us? Will they treat us well? Will they enslave us?
As a strong proponent of robots and AI, such questions also weigh on Dr Hanson’s mind.
“How can we make these machines benevolent, safe and caring?…We need to raise them among a human family and surround them with love…We must achieve super benevolent super intelligence,” he said.
The hope is that these super intelligence robots will choose the beneficial part of its learnings (and not be corrupted, like last year’s Microsoft’s AI chatbot experiment)
Humanity may well be heading down the road that blurs reality and fiction, till they are one and the same. If the future is as bright as Dr Hanson’s optimism, exciting times could be in store for all of us.
But will robots become truly human? Will they genuinely feel the deep longings of love, and the heavy grief of a passing?
Will they process these emotions into a bunch of algorithms, or like, as we humans do, relegate it to the mysteries of the heart and the mind? Will they really understand what it means to feel alive and fear death? Will they age? Will they yearn for an afterlife?
Those are the many questions humanity will have to confront in time to come.
—
The post A true living AI is adaptive, conscious, caring and ethical: Dr David Hanson of Hanson Robotics appeared first on e27.