Bernie Sanders, a senator and member of Congress from the U.S. state of Vermont, said he supports advancing artificial intelligence in healthcare and cancer treatment but that development efforts should be slowed because of the risks it poses to humanity.
Sanders said that even if the possibility of artificial intelligence producing devastating consequences for humanity is low, it cannot be ignored. He called for a more cautious approach because executives in the technology sector cannot provide definitive assurances about the risks and scientists have pointed to the possibility of a catastrophe for humanity.
Sanders said he wants artificial intelligence to be advanced in healthcare and cancer treatment, but added that he absolutely does not want the technology to develop into a process that could lead to the extinction of humanity. Referring to the 10 percent probability cited by Geoffrey Hinton, Sanders said the consequences of 5 percent or 3 percent probabilities also needed to be examined if they were to materialize.
Sanders to announce bill providing for a public stake
On June 3, Sanders said that the knowledge accumulated by humanity was being used without permission to create artificial intelligence models. In the same statement, Sanders announced that he would introduce a bill requiring companies in the artificial intelligence sector to give the public a stake.
Hinton says the consequences of artificial intelligence cannot be predicted
Nobel Prize-winning scientist Geoffrey Hinton said that humans had never before developed a product that could be more intelligent than themselves, making the consequences of artificial intelligence’s development impossible to predict.
Hinton said he did not consider it unreasonable to assess that the probability of artificial intelligence technologies causing the deaths of all humans within 10 years could be more than 10 percent. He said advanced artificial intelligence would not need to control physical systems to pose a serious threat and could also cause chaos by communicating with people.
Hinton also warned that high-capacity systems could be used in biological or cyberattacks and could contribute to the development of computer viruses. He stressed that artificial intelligence should be designed to prevent it from developing goals that would harm humans.
