Google has partnered with Oxford academic Stuart Armstrong to ensure their new artificial intelligence program, ‘DeepMind’ can be turned off by humans.
Google had been concerned the program, which relies upon a neural network style of computing that mimics human intelligence, might have been able to refuse to obey instructions, and use its intelligence to circumvent human authority.
Students held a variety of views on this collaboration and the possibility that technology may become so advanced that we cannot completely control it.
“It doesn’t seem at all likely to me that software would attempt to harm humans in order to keep its execution running”, said Magdalen JCR computer rep Winston Wright, who is interning with Google this summer.
On another level, the possibilities with artificial intelligence could greatly increase quality of life. “AI and machine learning have the potential to greatly help people do a variety of tasks. Areas that particularly excite me are automated medical diagnosis, driverless vehicles, and natural language processing,” Wright said.