“Artificial Intelligence can be interpreted in various ways, not only in comparison to human intelligence. The Turing test was primarily designed to measure humans in response to conversations, but now we sometimes call robots the ‘arms and legs of the internet of things.’ Basically, robots change the state of the world, they do not just spit out answers to a question, they actually change the state of the world. They interact with other humans and robots in the environment….So it’s more to do with whether we are able to we predict the actions that a robot takes, than whether a robot is behaving like a human or not.
…We live in a society where most of the time we are able to predict what another person is likely to do, and that is very crucial for the normal working of the society. Are we able to reliably predict the next set of action of robots, so that we can anticipate and utilise that set of behaviour? The risk is not that the robot may have some powers or behaviour that we can’t predict, it’s more like: do we understand the limitations of these robotic platforms?”
So the Turing test is irrelevant now?
Yes and no. We are involved in the creation of an anthropomorphic humanoid robot together with NASA for use on Mars. This is designed to go on unmanned missions to Mars to work on things like pre-deployment of assets on the red planet. I get asked the question: why the human form factor, why bipedal and not with ten arms? The answer to that question is that we have grown up in this human body and are used to working with it, which is similar in a way to the criteria under the Turing test. If we want seamless interaction to happen, then that is much easier if the robot takes the form of humanoid, that it has human like body structures, imitating what a human does but at distance of 10,000 light years away. This is a challenge and this is where the Turing test is still relevant.
Are robots a threat to human in terms of job replacements?
We have robots helping in areas like robotic surgery, in body part replacement, but more important, we are creating systems and capabilities that do not exist now. For example mining under the sea, on meteors, using robots to secure food security. My argument is that robots allow us to do things we could not do before or increase the goodness criteria of jobs which are already performed. Of course there will be a shift in the jobs that we do, there will be a shift in the demographics, and there is a need for a public debate about what is the next generation of skills sets we need, but that does not mean that robots are going to create mass unemployment. On the contrary, robots will help humanity need upcoming challenges.
Is Universal Basic Income a timely good development?
…UBI may provide the time and space needed to retrain, to bridge a period when some of the more mundane jobs are replaced. It may be a good idea to put some of the profits gained by robotisation into retraining people. If we we do this in the right way, people will lead a much more fulfilled life, because people will not think when they get up in the morning that they are a part of a big machine, when they go and pack chocolates; a job that robots can do much more easily. It is better to think of robots as things that will work in collaboration with his. There are still lots of skills which cannot be robotised.
…It’s up to us how we use the technology, in general, robots are our friends not our enemies.