
Analysis of Americans with $30 million or more in net worth from 2015-2025 challenges the dominant “California exodus” narrative. The data tells a more nuanced story about wealth concentration, AI’s geographic impact, and which states are actually winning and losing.
The Actual Numbers
California: 28,650 ultra-wealthy individuals, up 140% from 11,940. Market share actually increased from 14.3% to 14.9%. Silicon Valley’s AI boom created over 50 new AI billionaires in 2025 alone.
Texas: 18,720 ultra-wealthy (145% growth), the second-largest concentration—but still 35% smaller than California despite years of migration headlines.
New York: 16,560 ultra-wealthy, with share declining from 9.1% to 8.6%. Now surpassed by Texas in absolute numbers.
Florida: 11,240 ultra-wealthy with 185% growth—the largest percentage increase, validating tax migration narratives.
Illinois: 7,430 ultra-wealthy, share falling from 4.7% to 3.9%—the only major state showing meaningful relative decline.
The Mental Model Error
The “California exodus” story commits a classic mental model error: confusing anecdotes with data. Yes, some wealthy Californians moved to Texas and Florida. But AI wealth creation in California more than offset departures.
This illustrates second-order thinking in practice: while people focused on who was leaving, they missed the more important story of who was being created. Fifty new AI billionaires in a single year changes the calculus entirely.
The Real Story
Florida’s rise is real—tax policy works at the margin. Illinois’s decline is real—and uniquely severe among major states. But California’s position as the wealth creation engine remains unassailable as long as it dominates AI development.
The geographic distribution of wealth follows the geographic distribution of value creation. Currently, that means network effects in AI cluster around existing AI centers, reinforcing California’s position despite policy disadvantages. Disrupting that concentration requires either breaking the network effects or creating entirely new value creation centers.
Neither has happened yet.









