Erik Brynjolfsson is among the world’s most cited economists in information systems. A Stanford professor, NBER research associate, and Stanford Digital Economy Lab director, his work on how digital technology reshapes markets, competition, and wealth creation has proven consistently prescient. His central finding — that digital economics creates winner-take-all dynamics of unprecedented scale — has direct and urgent implications for how capital should be allocated in the AI era.

Why Digital Markets Concentrate at the Top

Brynjolfsson’s framework identifies the core economic property that distinguishes digital businesses from all that came before: near-zero marginal cost of reproduction. A software product, a model, a platform — once built, can be replicated and distributed to an additional user at essentially no cost. This is not true of physical goods, professional services, or labour-intensive businesses.

The consequence is a structural tendency toward superstar concentration. In traditional markets, competition is constrained by the cost of scaling. In digital markets, the best product can capture nearly all demand with minimal incremental cost. The result, as Brynjolfsson documents across industry after industry, is that a small number of firms capture a disproportionate share of total market value — while the rest compete for what remains. This is not a temporary imbalance. It is the natural equilibrium of digital economics.

‍“Today’s information technologies favour capital owners over labour, and increase the advantages that superstars have over everybody else. Future technologies will tend to increase spread, just as they will boost the bounty.”

Erik Brynjolfsson — Stanford Digital Economy Lab, co-author ‘The Second Machine Age’ ‍

‍

AI Accelerates the Superstar Dynamic

Generative AI intensifies every element of the superstar dynamic Brynjolfsson identified in the digital economy. The frontier model companies — those with the compute, the data, and the talent to train at scale — are not merely building better products. They are building reinforcing advantages: more users generate more data, which trains better models, which attract more users. The positive feedback loop is structural.

Brynjolfsson’s concept of the ‘productivity J-curve’ is also instructive: general-purpose technologies like AI take time to translate into measured productivity gains, as complementary organisational changes lag technology adoption. We are currently in that lag phase — which means the full economic impact of today’s frontier AI companies is not yet visible in GDP statistics, but is already visible in the compounding of their strategic position.

Owning the Superstars Before the Market Does

The challenge for investors is that the companies at the apex of the AI superstar hierarchy are private. Anthropic, OpenAI, SpaceX, Lambda — these are not available through any index, ETF, or public equity strategy. By the time they list, public markets will have fully priced the superstar premium. The opportunity to acquire ownership at pre-public valuations belongs exclusively to those with access to private secondary markets.

The Exto Global Technology Leaders Fund is built around this insight. Through secondary market transactions in proven, late-stage AI and technology companies, the Fund provides qualified investors with ownership stakes in the businesses most likely to define the superstar landscape of the coming decade. Brynjolfsson’s work identifies these companies not as speculative bets, but as the natural beneficiaries of digital economics at scale — and the secondary market provides the means to participate on terms that public markets will never offer again.

‍

‍