Well, This Is Awkward: Economists Finally Admit AI Might Actually Replace Us (Sorry, “Vibes” Didn’t Hold Up)

Remember when economists were the designated chill squad every time someone panicked about robots taking jobs? “Relax,” they’d say, sipping overpriced oat milk lattes. “ATMs didn’t kill bank tellers. Excel didn’t wipe out accountants. And Roombas? Bless their little spinning hearts, they still can’t handle dog hair under the couch.” The consensus was basically: automation anxiety is just a mood — don’t let the bots ruin your vibe.

That vibe just got yeeted into the void. A fresh paper — co-authored by brain trusts from the Fed, Yale, Stanford, and a few other institutions that probably still think PowerPoint is cutting-edge — surveyed 69 economists, 52 AI nerds, and 38 superforecasters (yes, that’s a real job now). The shocking twist? All three groups actually agree on something. Cue the apocalyptic music: faster AI progress means fewer humans working. Not “working less.” Working. Period. The labor force participation rate? Yeah, it’s going down — like a DeFi token post-airdrop.

Under the “rapid” scenario — where AI outsmarts humans across most cognitive and physical tasks by 2030 — economists predict the U.S. participation rate will nosedive from 62% to 54% by 2050. That’s roughly 10 million jobs evaporating directly because of AI, not because boomers are retiring or Gen Z is too busy making TikTok ASMR. This isn’t “AI helps us work better.” This is “AI files your taxes, defends you in court, and then writes a Yelp review about how inefficient you were.”

And let’s be clear: the rapid scenario isn’t Black Mirror fan fiction. It’s the world where AI negotiates book deals, runs factories, debugs your smart home, and makes freelance devs cry into their ergonomic keyboards. Even Anthropic’s Dario Amodei — a guy who literally builds AI gods — has warned this is moving faster than anyone expected. So when economists nod along to that timeline, it’s less “interesting hypothesis,” more “we’re already behind.”

GDP, of course, is out here throwing a pool party. Under the same doomscape, economists project annual growth of 3.5% by 2045–2049 — a number that hasn’t been seen since the post-WWII economic jolt. AI experts, never ones to underhype, are calling 5.3%. That’s a lot of wealth being created — just don’t expect a slice unless you’re already in the 0.1%. It’s like the entire economy is running on a single ETH whale’s portfolio.

The paper casually drops that under rapid AI, the top 10% could own 80% of total wealth by 2050. That’s not just Gilded Age levels — that’s “feudalism with better Wi-Fi.” The American Dream is now a liquidity pool with diminishing returns.

But here’s the part that gets lost in the doomscroll: the real debate isn’t whether superintelligent AI is coming. It’s what happens after it arrives. The old techno-optimist fairytale was that automation always creates new jobs — like how elevators needed operators (RIP). But AI might not just replace workers; it might automate the very act of inventing new jobs. It’s like if the guy who came up with “influencer” was himself an AI. We’re not just out of work — we’re out of ideas.

For now, the overall employment numbers look… fine? A Yale and Brookings study from late 2025 found no mass unemployment wave nearly three years after ChatGPT dropped like a viral NFT. But zoom in, and the cracks show: workers aged 22–25 in AI-exposed roles have already seen a 13% relative job drop. So macro? Stable. Micro? Looks like a rug pull.

On policy,