Daron Acemoglu is not an AI optimist. The MIT Nobel laureate has spent his career studying how technology reshapes power and prosperity — and his findings on AI are a sobering counterpoint to industry hype. But for the sophisticated investor, his conclusions carry a message that is, paradoxically, highly actionable.

The Capital–Labour Divide Is Already Widening

Acemoglu’s landmark paper, ‘The Simple Macroeconomics of AI’, published in Economic Policy, establishes a clear finding: AI advances are not reducing labour income inequality. They are widening the gap between capital income and labour income. The mechanism is structural. As AI automates tasks previously performed by workers, value creation shifts from labour toward the owners of the capital assets driving that automation.

His modelling is deliberate in tempering broad productivity expectations — GDP gains over the next decade will be real but modest. What is not modest is the distributional effect. The productivity gains that do materialise will accrue disproportionately to capital owners, not workers. Wages at the median are unlikely to keep pace with returns to capital in AI-intensive sectors.

‍“AI is predicted to widen the gap between capital and labour income. If left to market forces, the benefits would accrue primarily to those who own capital and control AI technologies.”

Daron Acemoglu — Nobel Laureate in Economics, MIT Institute Professor

What This Means for Portfolio Thinking

Acemoglu’s work points to a durable structural shift, not a cyclical one. If the labour share of income is in long-run decline — and the capital share is rising — then traditional portfolio construction anchored in labour-correlated returns faces a structural headwind. Government bonds, consumer income plays, and businesses dependent on wage-driven consumer spending all carry embedded exposure to this trend.

The inverse is equally clear: direct ownership of AI-enabling capital — the compute infrastructure, the frontier model companies, the platforms generating AI-driven productivity — is precisely the asset class positioned to capture the value that labour is losing. This is not a sector bet. It is a positioning decision about which side of the most consequential economic divide of our time you are on.

The Access Problem — and the Exto Solution

Acemoglu himself notes that concentration in AI ownership is already a reality — wealth inequality is “already happening due to hype driving valuations of companies in a concentrated industry through the roof.” The implication for investors is that delay compounds the problem: the concentrated nature of AI capital means the window for meaningful ownership at sensible valuations narrows over time.

The Exto Global Technology Leaders Fund provides qualified Australian investors with direct secondary market access to the companies at the apex of this concentration. Rather than waiting for these businesses to list publicly — by which point public markets will have priced in the structural advantage — the Fund enables ownership now, at valuations anchored to today’s fundamentals. This is capital deployment that aligns directly with the economic logic Acemoglu has spent a career establishing.

‍

‍

‍