Self-aware robots with deadly intentions are a staple in sci-fi films. Now scientists have proved a robot can show a glimmer of consciousness – but luckily this android is polite. In an experiment a small humanoid solved a philosophical problem to demonstrate it could understand a question and then recognise its own voice.
Scientists at The Rensselaer AI and Reasoning Lab (Rair) in New York conducted an experiment on three Nao robots. Told the bots two of them had been given 'dumbing pills' to silence them. When asked which had been given the pills, they all tried to respond, but one spoke and recognised its own voice to solve the problem. Scientists claim this shows the robot demonstrated self awareness
Scientists have proved a robot can show a glimmer of consciousness. In an experiment, Nao bots were programmed to think two of them took 'dumbing pills'. They all tried to respond to a question about which robots took the pills, with one (right) able to recognise its own voice and reason it could speak
Scientists at The Rensselaer AI and Reasoning Lab (Rair) in New York, led by Professor Selmer Bringsjord, conducted an experiment on three Nao robots, which are commonly used in research projects.
They were programmed to think two of the group had been given ‘dumbing pills’ to prevent them from speaking, while one had been given a placebo, PopSci reported.
In practice, this meant that mute buttons were pushed on two robots, leaving one able to speak.
If this research is anything to go by, it may be that the first fully conscious bots will be more like Star Wars’ C-3PO (pictured left) than The Terminator (right)
But the robots didn’t know which one of them had retained the ability to speak.
The group of androids was asked which two had been given the pills.
While all of them tried to answer, only one was able to respond aloud by saying ‘I don’t know’.
It seemed to take a while to consider the question, before standing up and answering.
The robot that spoke recognised its own voice and said: ‘Sorry, I know now.
'I was able to prove that I was not given a dumbing pill.’
Its response demonstrates a basic level of self-awareness ... as well as good manners.
The test also shows that the robot could understand a question as well as demonstrate reasoning and recognise its own voice.
If this research is anything to go by, it may be that the first fully conscious bots will be more like Star Wars’ C-3PO than The Terminator.
Professor Bringsjord will present the findings of his research at RO-MAN, a robotics and artificial consciousness conference in Japan that begins in late August.
Dr Stuart Armstrong predicts that in the future, intelligent robots will be smarter and faster than humans, with the ability to take over the running of countries (illustrated with a stock image) and even wipe us out altogether
MACHINE CONSCIOUSNESS VERSUS INTELLIGENCE
Machine consciousness differs from machine intelligence.
There are already intelligent machines that can learn from their mistakes and fix errors, but many of these are programmed to do so.
For example, world chess Grandmaster, Gary Kasparov, was defeated by IBM’s ‘Deep Blue’ supercomputer in 1997.
Apple Siri, Google Photo Search and Google’s self-driving cars are further examples of machines that exhibit behaviour similar to that exhibited by intelligent humans.
To become conscious, Andrew Sheehy from Generator Research said systems would need to be able to look at the rules the humans had programmed, to identify inconsistencies or gaps and report back.
Alternatively, they would be able to discover errors in algorithms and consciously fix them without being asked.