AI's Slow Tide: Job Obsolescence Fear Calibrated

A significant MIT study, "Crashing Waves vs. Rising Tides," re-frames the fear of becoming obsolete (FOBO) concerning AI. Researchers analyzed over 17,000 evaluations of AI outputs across 3,000 labor market tasks, involving more than 40 AI models. The findings indicate AI's progress is more akin to a "rising tide," showing steady, simultaneous gains across many tasks rather than abrupt disruptions. Today, frontier AI models achieve a 50% success rate on tasks requiring a full workday for humans, with failure rates halving roughly every two to three years.

While AI's capabilities are advancing rapidly, with projections suggesting 80%-95% success rates on text-based tasks by 2029, most companies lag in adoption, with fewer than 19% of U.S. establishments having integrated AI. This organizational inertia, coupled with inadequate worker training – down nearly 10 percentage points since 2024 – exacerbates FOBO. Companies actively deploying AI are already experiencing substantial productivity gains, recapturing 40-60 minutes per worker daily. Legal work shows the lowest AI success rates at 47%, while installation and repair tasks reach 73%.

Despite AI's expanding efficacy, significant hurdles like integration costs and liability concerns remain. The MIT study emphasizes that widespread automation, especially in error-intolerant fields, may still be some distance away. Workers are advised to treat AI as a tool for adaptation, as the "rising tide" offers a window for adjustment rather than immediate displacement. The fear of obsolescence is rational, but paralysis is counterproductive.