You’ve heard the warnings. Now they’re coming with spreadsheets.
Anthropic CEO Dario Amodei didn’t mince words this week: generative AI could wipe out 50% of entry-level white-collar jobs within five years. His prediction? Up to 20% unemployment if no guardrails are put in place.
Harsh? Sure. But think about it. If an LLM can summarize earnings calls, write product descriptions, or sort customer complaints faster and cheaper than a human, why wouldn’t a company take the shortcut?
The threat isn’t theoretical anymore—it’s operational. Law firms are already using GPT-powered tools for document review. Marketing agencies use AI to spin out ad variants by the dozen. And customer service? That’s practically run by bots already.
You don’t need a crystal ball. You need a quarterly report. And those reports are pointing to reduced headcounts in departments where AI can fill the gaps.
But here’s the rub: the AI still needs us. It mimics, regurgitates, and blends what we’ve already built. Creative prompts, strategic decisions, and judgment—that’s the human edge. At least, for now. AI can draft 20 LinkedIn posts, but it doesn’t know which one will get someone hired to do a white-collar job. Not really.
So what’s next? We’ll likely see a redefinition of work itself. Entry-level roles may shift into AI-assisted roles. Think “prompt engineer” or “automation analyst.” Mid-career professionals might become reviewers and curators rather than originators. Training will matter. So will adaptability.
The bigger question isn’t what AI can do. It’s whether we’re ready for what happens when it does it better, faster, and cheaper. Are we planning for that? Or just hoping we can prompt our way out of it?