Several pages on this website have addressed the idea that systems, programs and devices frequently described as AI are no such thing. They are clever, well-programmed computers but intelligent? Gratifyingly, a conference sponsored by the White House Office of Science and Technology Policy has come to the same conclusion.
The reason for the conference was to discuss whether government oversight of AI was required to prevent the technology from causing social chaos by throwing too many people out of work. The conclusion: the human mind has a level of flexibility and a learning capability that AI is still a long way from matching. Autonomous systems – that is, systems that have been programmed to examine data and make decisions without human input – are already making decisions in health, transportation and warfare. They are not, it would seem, in any danger of pushing people out of the more creative occupations for some time yet.
As one computer scientist at the Office of Science and Technology Policy put it, the AI community climbs mountain after mountain, but all it sees after that is more mountains.
Genetic engineering is still seen as a far more potent threat to the human race than artificial intelligence. While Hollywood has made great play (and lots of money) from AI, for example in the Terminator movies in which the computer network Skynet becomes an enemy of humanity, such ideas remain in the realm of dreams and sci-fi fantasy.
Algorithms guiding missiles are useful but are not intelligent – they depend on the activities of humans. The same can be said of smart phone services and many other things routinely described as AI. The Hello Barbie doll speaks and listens. It hears the child and feeds what it hears back to an enormous database processor. While that may make it come up with new phrases that enchant the child, it is perhaps time for parents to be more concerned about who is listening to their child and what may ultimately be done with the information being harvested.