Artificial General Intelligence Likely to Outperform Humans in All Tasks within 45 Years
Written by Stephan Cunningham, vice president, product management at Veritone
Young people considering a career in retail, translation or truck driving may want to rethink their options.
That’s because artificial general intelligence is poised to surpass human capabilities in these jobs and a broad range of other vocations during the coming years, according to a survey of AI experts conducted by the Future of Humanity Institute. The survey respondents predicted artificial general intelligence has a 50 percent chance of outperforming people in all tasks in 45 years and automating all human jobs in 120 years.
While the survey anticipates artificial general intelligence will have a positive impact on society, the experts also called for more research on how to minimize the risks the technology could represent for human civilization.
“Advances in AI will have massive social consequences,” the survey report noted. “Self-driving technology might replace millions of driving jobs over the coming decade. In addition to possible unemployment, the transition will bring new challenges, such as rebuilding infrastructure, protecting vehicle cyber-security, and adapting laws and regulations.”
Researchers predict AI will outperform humans in language translation by 2024. By 2026, artificial general intelligence will be able to write a high-school level essay better than a person. AI will surpass human truck drivers by 2027.
In the longer term, AI will outpace humans in a range of other jobs, including retail in 2031, writing a bestselling book in 2049 and performing surgery in 2053.
While the displacement of human labor is a major concern, it’s hardly the worst outcome foreseen by some of the survey respondents. The survey put a 5 percent probability on an “extremely bad” outcome from the rise of AI. In case you are fuzzy on what “bad” means, the report mentions “human extinction” as an example.
This negative outlook reflects warnings from top thinkers about the ramifications of developing human-level AI. Among those sounding the alarm are physicist Stephen Hawking, who predicted the development of full artificial intelligence “could spell the end of the human race.”
However, there’s a much higher chance of a good outcome, with a 45 percent chance of positive results arising from the deployment of artificial general intelligence.
While many of the forecasts in the survey represent speculative predictions of events far in the future, the provenance of the study endows it some credibility. The Future of Humanity Institute is a research organization based at the prestigious University of Oxford. The survey was conducted in cooperation with researchers from Yale University.
Stephan Cunningham is vice president, product management at Veritone. Working in concert with core internal teams including industry-specific general managers and engineering as well as directly with clients and prospects, he leads the disciplines and business processes which govern the Veritone Platform.