Here's why waiting for AI to settle down is the expensive move

Someone you love asks whether they should finally go for it with AI, and you answer in four seconds. Your own version of that question has been sitting unanswered for a year. This one's about why the sceptical headlines about companies failing at AI were never actually the evidence that mattered for you — and what the last time a skill premium like this closed can tell you about how much time is actually left.

In this memo

1. 0:00 The four-second answer
2. 0:49 The premium everyone's fighting over
3. 2:36 What the spreadsheet actually measured
4. 4:43 Why the window closes quietly
5. 7:22 Put a number on staying

Sources

• PwC, 29th Annual Global CEO Survey (2026)
• PwC, 2026 Global AI Jobs Barometer
• WRITER / Workplace Intelligence, 2026 AI adoption survey
• MIT NANDA, "The GenAI Divide: State of AI in Business 2025" (preliminary, not peer-reviewed)
• Lee & Kim, "How Information Technology Affects Wages," Journal of Labor Research (2002)
• BLS travel agent employment data, via Stripe Economics
• Akerlof, "Procrastination and Obedience," American Economic Review (1991)

Watch on YouTube ↗ Subscribe ↗