AI Likely to Outperform Humans in Less Than 50 Years
If you think your job is safe from the reaches of artificial intelligence, it's time to think again. A new survey of experts finds that they believe AI can overtake humans in most tasks within the next 50 years.
In a new paper "When Will AI Exceed Human Performance? Evidence from AI Experts," researchers from Yale and Oxford universities conclude that there is at least a 50% chance that in the next 45 years, AI will perform most tasks better than the human who created the technology.
And in 120 years, AI will perform all human jobs, predict the experts.
It's not just redundant or mundane tasks, such as those in call centers, that AI will take over. Instead the paper finds that the technology will drastically alter transportation, healthcare, the military, finance and science, and that the time for a robust policy discussion is now.
"New challenges, both for AI developers and policy-makers, will also arise from applications in law enforcement, military technology, and marketing. To prepare for these challenges, accurate forecasting of transformative AI would be invaluable," the researchers wrote.
The report is based on responses to questions sent to 352 researchers in the field of high-level machine learning or HLML. Each was asked to estimate when he or she thought HLML or AI would arrive and take over most human tasks.
About 50% responded that it would be in 45 years, with a minority --10% -- reporting that HLML would arrive in nine years.
If you are wondering when your job will be performed by a machine, here's a sample:
- Translating languages by 2024
- Writing high school essays by 2026
- Driving a truck by 2027
- Retail work by 2031
- Writing a best-selling book by 2049
- Working as a surgeon by 2053
While it's easy to get into discussion about when Skynet will take over or if humans will live lives like in Wall-E, this report does dovetail with a recent Gartner report that found AI will not only take over low-skilled jobs, but high-skilled ones as well, including medicine, the law and information technology. (See AI Is Coming for Highly Skilled Jobs, Too.)
At the heart of these studies is a significant problem: It's not a matter of when AI will start taking over human job, but how should we plan for this future? (See Will AI Create More Jobs Than It Destroys?)
"In addition to possible unemployment, the transition will bring new challenges, such as rebuilding infrastructure, protecting vehicle cyber-security, and adapting laws and regulations," according to the Yale and Oxford report.
Still, there is room for some optimism. The Yale and Oxford researchers found 25% of respondents believe that HLML and AI will have a "good" outcome, while 20% believe it will have a "very good" outcome. In contrast, about 10% responded these technologies would have a "bad" outcome, and only 5% described it as "extremely bad," which means human extinction.Related posts:
- Google Placing Multiple Bets on AI – Reports
- Cisco Buying AI Specialist MindMeld for $125M
- AI, Cognitive Spending Soaring to $12.5B in 2017