ARK Invest's Chief Futurist Brett Winton argues AI will substitute for knowledge work rather than just assist it, putting a roughly $30 trillion global wage bill in play. His firm projects AI-driven software spending could reach $3 trillion to $7 trillion a year by 2030, with frontier model companies capturing the largest share of that value.
Winton frames the opportunity around the world's knowledge workers — accountants, analysts, lawyers, consultants, coders, and marketers — whose collective wage bill he pegs at roughly $30 trillion, excluding China. His thesis holds that AI will not simply speed up that work; it will begin replacing it, redirecting labor spending toward software and automation.
The numbers behind the thesis
ARK projects that AI-driven software spending could reach $3 trillion to $7 trillion annually by 2030, against a global software market currently measured in the low single digits of trillions. The firm is arguing AI alone could double or triple that market within half a decade.
Underpinning the forecast is a cost curve: AI inference costs on agentic benchmarks have reportedly been declining more than 99% year over year, a pace ARK expects to hold at least through mid-2026.
Who captures the value
Winton sees frontier model companies as the primary beneficiaries, with ARK estimating their potential enterprise value at $15 trillion to $20 trillion or more. OpenAI's annual recurring revenue is reportedly in the $20 billion to $30 billion range, while Anthropic has been growing even faster than its competitors in recent quarters.
ARK also estimates that compute capacity rents at roughly $10 billion per gigawatt and can monetize output at around $30 billion per gigawatt — meaning every dollar spent building and powering data centers could return roughly three dollars in economic value.
A 7% GDP growth case
ARK forecasts that AI could accelerate annual GDP growth to 7%, well above most mainstream estimates; the US economy has averaged somewhere around 2-3% real growth in recent decades. For investors, the key variables to watch are adoption speed and margin compression: faster enterprise adoption pushes toward the $7 trillion end of ARK's spending range, but if competition drives model prices toward zero, the $15 trillion to $20 trillion enterprise-value estimate could prove too generous, shifting value instead to the application layer and end users.
The $30 trillion wage bill will not disappear overnight. But redirecting even 10-20% of it toward software and AI services over the next decade would mark one of the largest wealth transfers in economic history.
Source: Crypto Briefing
Trading involves risk.
Risk Disclaimer
