AI Is Actually Taking Jobs, Starting With Claims Adjusters
The 'You Know Who Hates AI' thread (49508225) cites BLS data showing insurance claims adjusters lost 21% of jobs between May 2025 and May 2026. This isn't speculation about future displacement; it already happened. The thread debate is whether AI is better at the job or just faster at making bad decisions at scale, with several commenters arguing that speed and cost matter more than accuracy to insurers, which is its own kind of damning.
The pattern here is that white-collar, rules-heavy work with large document volumes is getting automated first. Claims adjusting fits that profile perfectly: lots of text, clear decision criteria, high volume, and an industry with strong incentive to cut costs. The question isn't whether AI can do the job perfectly, it's whether it can do it cheaply enough that perfection stops being the standard.
This connects to the broader AI skepticism thread (49526069) where Ed Zitron is criticized for letting his skepticism become a political identity rather than an analytical position. The irony is that the jobs data he might have cited to be bearish is now coming in bullish for displacement.
So what?
If you're building AI tools for industries with high-volume, rules-based decision workflows, the market validation is arriving in real-time. But founders should note that the displacement is happening because AI is cheap and fast, not necessarily because it's better, which means the regulatory and liability backlash is building alongside the adoption.