Computers should be trained to serve humans to reduce their threat to the human race, says a leading expert on artificial intelligence
We are all doomed to a dystopian future run and controlled by smart machines of our own making - or perhaps not.
At a session at the World Econoimc Forum in Davos, Stuart Russell, a leading expert on artificial intelligence (AI) and robotics, made the bold prediction that AI would overtake humans “within my children's lifetime”.
The chief challenge was to control these advances by making sure that computers continue to serve human needs, rather than become a threat to them, the Berkeley professor argued.
To do so, it was imperative that robots were endowed with the same values as humans.
Professor Russell defined the ideal relationship as similar to that of Bertie Wooster and Jeeves where the long-suffering butler understands perfectly what his master wants without needing to be told.
However, it was possible to envisage a much more dangerous future where “sociopathic” robots become a threat to humans.
The biggest immediate danger from artificial intelligence was autonomous weaponry, followed in the medium term by disruption to established forms of employment. Computers have already rendered redundant millions of once well-paying white collar jobs.
In a recent open letter signed by Professor Russell, and among others, the cosmologist Stephen Hawking, it was argued that urgent research was needed into how to limit the destructive potential of AI.
“The potential benefits are huge, since everything that civilization has to offer is a product of human intelligence; we cannot predict what we might achieve when this intelligence is magnified by the tools AI may provide, but the eradication of disease and poverty are not unfathomable. Because of the great potential of AI, it is important to research how to reap its benefits while avoiding potential pitfalls”, the letter said.
Others were more sanguine. “Humans constantly learn from their mistakes. We have a built in brain system that allows us to make continuous adjustment, a process which cannot be replicated by decimal point precision”, said Robert Thomas Knight, professor of psychology and neuroscience at Berkeley.
His Berkeley colleague, Alison Gopnik, thought human stupidity would always be a much greater threat to our future than AI.