The push to swap coders for AI is costing Big Tech more than it saves, which is bad for junior developers, investors in the biggest tech stocks, and eventually the wider economy.

• AI writes code much faster, but twice as much of it now gets thrown out or rewritten, and close to half of it carries a known security hole.
• Companies that adopted AI tools heavily barely shipped more work, while the true cost of owning AI-written software is heading 30-40% higher than traditional development.
• Junior hiring has collapsed while demand for senior engineers jumps, turning experienced developers into expensive cleanup crews and breaking the pipeline that creates future seniors.
• The bigger driver is money: the five biggest tech firms carry $1.65 trillion in off-balance-sheet obligations from data centers and hardware deals, run up when borrowing was cheap and now costly to service.
• "AI efficiency" has become the preferred cover story for layoffs because markets reward that framing, a habit now widely called AI washing.

Outlook: The most likely path is a slow deflation of the hype rather than a crash, with years of cost-cutting, a permanently thinned developer workforce, and a real but smaller risk that a few big defaults trigger a credit crunch.