According to MIT’s statistical models, AI could already replace 11.7% of the U.S. workforce. By that, they don’t mean theoretically, in the distant future. They’re saying that 11.7 percent of the US workforce can be replaced by AI right now.
That finding comes from a sprawling research project called The Iceberg Index, which is equal parts economic forecast and sci-fi speculation. The whole thing is an effort to get a better sense of how AI will shape the future.
Behind it all is a massive simulation powered at Oak Ridge National Laboratory, which is affiliated with the Department of Energy. The director of AI programs at Oak Ridge explained the concept to CNBC by saying, “Basically, we are creating a digital twin for the US labor market.” Think of it like The Matrix but with the fidelity of a spreadsheet.
What This Means for the Future?
The simulation uses a “Large Population Model” that tracks the behavior of 151 million workers as if they were autonomous agents with 32,000 measurable skills. It even accounts for geography, because where you live will likely play a significant role in whether AI replaces your job.
Of course, the intended audience for all of this is lawmakers and CEOs, the latter of which and some of the former are likely champing at the bit to eliminate as many jobs as they can to make corporate bottom lines shine a little brighter at the expense of the rest of us. They’re told the tool can help them target “exposure hotspots,” tweak policy ideas before writing billion-dollar checks, and map out training needs.
There is a big asterisk attached to all of this. And if the same one you find in a lot of speculative science, whether it be related to technology, astrophysics, or health, correlation does not equal causation.
In other words, just because AI can theoretically take over 11.7 percent of US jobs right now doesn’t mean it will happen, and most importantly, it doesn’t mean it even can happen. A loss of 11.7 percent of jobs in an already ailing job market is disastrous, apocalyptic even.
That wouldn’t immediately equal Great Depression-era levels of unemployment. Still, it exceeds a Great Recession-era peak of 10 percent unemployment set in October 2009— and that’s not even counting the unemployment rate that would have preexisted before that even happened, so you have to add those numbers on top of it.
There’s also the matter of whether AI can even do any of these jobs well enough to take them permanently. There are a lot of ifs here. Still, the paper insists that policymakers shouldn’t wait for perfect, irrefutable evidence to act. Which direction they act—whether in favor of corporations or of people—remains to be seen.
The post MIT Study Says AI Is Ready to Take a Whole Lot of American Jobs Right Now appeared first on VICE.
