Are We on the Cusp of AI Turbulence? | American Enterprise Institute

When the stakes are high and time is short, those advocating interventionist economic policy have a stronger-than-usual case for urgent and direct action. During the COVID-19 pandemic, Washington worked with industry to speed vaccine creation and deployment. With thousands of Americans dying every day, time was of the essence.

Is the clock also ticking when it comes to artificial intelligence? In a nearly 6,000-word essay, Microsoft cofounder Bill Gates claims AI is improving at a “mind-blowing rate”—hardly an uncommon opinion among technologists and tech-company CEOs—and that global governments need to “act now” before they are overwhelmed by disruption that is educational, economic, and borderline existential. (He thinks AI can do lots of good stuff, too, it should be noted.)

Time is short, so time to get planning, tout de suite. Among Gates’ buzziest ideas are setting aside some jobs for humans only and taxing AI tokens and robots to give human workers some financial edge over the machines.

(The superbillionaire adds that “if someone had a credible plan for slowing down AI advances globally, I would likely support it. However, I don’t think that’s going to happen. The geopolitical and economic incentives are pushing too hard to go full speed ahead.”) 

The piece is titled “The Turbulent AI Era Is Here. The choices we make now are critical,” but I’m not sure about the indicated timeline. Neither is OpenAI boss Sam Altman, who in a podcast the other day conceded he had been overly optimistic about the pace and breadth of productive AI adoption, underestimating how the economy “just has so much inertia.” It’s an assessment backed up by considerable ongoing research, such as a recent Federal Reserve staff analysis that finds the economy still in the AI “buildout phase” rather than one of “broad-based displacement.” Likewise, Goldman Sachs finds AI’s job market impact to be ”visible but narrow.”

Still, maybe the American economy is on the cusp of rapid transformation. It’s impossible to know for sure. Yet however impressive AI models are in the lab, all manner of real-world constraints and bottlenecks remain, from energy to business reorganization to public approval.

I also can’t help thinking about the story of radiologists. A decade ago, Nobel laureate Geoffrey Hinton, the “godfather of AI,” basically warned people to stay away from that profession due to computerized competition. Back then, radiology might have seemed like a job meriting preservation for humans.

But a new Ars Technica piece notes that “radiology’s ranks are in fact growing steadily, with the number of practitioners expected to expand by 26 percent or more over the next three decades,” and the big challenge is optimizing collaboration between humans and computers. Maybe there will be more job augmentation versus automation than some AI worriers think.

Predictions are hard, it’s been said, especially about the future—which doesn’t mean ignoring the potential impacts of possible scenarios. But policymakers should maintain considerable optionality as we all watchfully wait for what hard data and market signals tell us about the emerging Age of AI.