Artificial Intelligence, Markets, and the Difference Between Short-Term Winners and Long-Term Value

AI's long-term value may accrue less to today's obvious infrastructure winners than to trusted incumbent businesses that use AI to improve economics, while valuation remains decisive.

• Semiconductor and data-centre infrastructure companies have captured the market's attention, but passive flows into trillion-dollar companies can create reflexive price increases and embed increasingly aggressive expectations.
• Micron Technology illustrates the risk: EPS was approximately $7.50 in August 2025, while analysts were contemplating EPS 20 times higher 12 months later as AI-memory demand surged; subsequent sentiment shifts produced a significant de-rating.
• S&P Global is an example of feared disruption strengthening an incumbent: it reported pricing premiums of roughly 35–45% on AI-ready datasets and enhanced analytical offerings during current contract renewals.
• Salesforce may also be capturing AI value through stronger demand, customer engagement and new revenue streams, so excluding AI products from its underlying growth analysis can omit an important part of its economics.
• B2B businesses may be more resilient than B2C alternatives because enterprise systems involve trust, continuity, integration costs and multiple human decision-makers; the article argues that AI can deepen these relationships, raise productivity and support higher-value services.

Takeaway: Investors should distinguish economic beneficiaries from stocks that have already benefited in the market, assess the payoff as well as the probability of disruption, and avoid paying excessive prices; Brown Advisory reports that AI concerns affect roughly 20% of the Global Value Select Strategy and 15% of the International Value Select Strategy, while warning that no thematic exposure should dominate a portfolio.