The Laboratory of Nuclear Studies, Cornell University, 1948. Photo: Bettmann/Getty
The ultimate goal of AI research is a system that rivals a human’s thinking power — but it is still far out of sight.
Instead, in the coming years, AI may provide an intelligence boost in a different way: by coordinating still-unsurpassed human brainpower and correcting some of the errors inherent how we think.
By definition, AI processes information very differently from humans and can combat group-think, says Berkeley’s Ken Goldberg, a robotics expert who directs the university’s People and Robots Initiative.
- One way to do this is by simply assembling better teams. Algorithms can create groups with complementary skills and qualities, each challenging their peers and filling in others' gaps — like an automated Tinder for work.
- IBM is developing technology meant to prompt critical thinking in humans, in the form of what Goldberg calls an AI-powered devil’s advocate.
- In a similar vein, bots will automate humans' better nature, reminding decision-makers about the context and ramifications of their choices, or of potentially false assumptions.
These advances would improve work for the creative class. But for workers whose time is taken up almost entirely by routine tasks, AI may be more threatening than liberating.
- In the long run, even if more jobs will be created than lost, many millions may need to find new work. The early ripples of this force are already moving through the job market, squeezing professions built around rote work.
- These jobs may not benefit directly from creativity-boosting AI.
Go deeper:
- Inside the quiet AI revolution (Axios)
- Goldberg's report, AI & The Future of Work