2 November 2018

Sorry, Sophia, but you're just a big fraud

Over the last year or two there has been an avalanche of hype in the media about intelligent robots. According to some artificial intelligence (AI) geeks, we are on the threshold of the "Singularity" - the point at which machine intelligence will surpass human intelligence and trigger runaway technological growth, resulting in unfathomable changes to human civilization.

If only.

The leader of the mechanical pack is a robot called Sophia who takes part in TV interviews and even participates in platform debates on the future of robotics. Sophia made headlines last year when "she" was granted citizenship of Saudi Arabia.

Sophia's creator, David Hanson, of Hanson Robotics, claims that his robot is self-aware, or "becoming self-aware", and that "she" understands emotions and recognises the people around "her".

Hanson is a liar and a fraud (and he's welcome to sue me if he'd like to contest that description in court).

It would be nice (potentially) if robots were intelligent, self-aware etc., but they aren't. The reality is that we're not even close to developing true artificial intelligence, and David Hanson is misleading the public - as well as scamming his investors - by claiming that his creation is self-aware.

Robots like Sophia (strictly speaking, Sophia is an automaton, not a robot) are programmed to respond to specific questions and keywords. This is achieved with a relatively simple text-to-speech synthesis system. There is zero intelligence involved. Interviews with journalists are scripted and choreographed in advance. The answers to specific questions are pre-programmed. And an "operator" has to be present off camera to ensure that Sophia is fed the right questions.

In other words Sophia is a fraud. A scam. A confidence trick to dupe wealthy but scientifically-illiterate sponsors into investing enormous sums of money into AI technology - and specifically into the bank accounts of Sophia's creators.

It looks impressive when Sophia gives "intelligent" answers to journalists' questions, accompanied by appropriate facial expressions, but it's all an illusion. If Sophia were asked a random question that "she" hadn't been programmed to answer, "her" only response would be: "I don't understand that question."

The same applies to all the other so-called intelligent robots out there. They're essentially no more advanced or "intelligent" than the clockwork automata that were around 200 years ago. The realistic humanoid appearance and facial expressions trick the mind of the observer into thinking that there is some kind of awareness going on in the robot's "mind". There isn't. And, unfortunately, all this talk of "singularity" is baloney. The truth is we are still a long, long way from anything approximating real artificial intelligence. Our best hope lies in the development of quantum flux computers, but that area of science is still in its infancy.

In the meantime, David Hanson is doing a disservice to science - and hoodwinking his supporters and patrons - with his fraudulent demonstrations of artificial intelligence and his grossly exaggerated claims. The media are also culpable for their uncritical coverage of this alleged technological breakthrough. In many cases it is evident that they are tacitly colluding with Hanson and his colleagues for the sake of entertaining the viewing audience and boosting their ratings.