Market participants are focused on whether artificial intelligence investments will yield a return on investment as compute infrastructure expands. The central question driving current valuations is which companies will successfully monetize these capabilities profitably.
The Market Hang panel featured Dan DeFrancesco, executive editor at Business Insider Today; Pras Subramanian from Yahoo Finance; Liz Harrow, a partner at SHAKTI; Turney Duff, a Wall Street veteran; and Brian Mulberry, chief market strategist at Zacks Investment Management. DeFrancesco argued that AI will touch every part of life, stating that the sector is not yet in the first inning of its transformation. He noted that non-tech companies will be dramatically different in three years and expressed a willingness to remain long despite potential declines of 15 to 30 percent.
Subramanian emphasized that the market's direction depends on who can monetize AI profitably. He observed that the S&P 500 trades at a high premium due to capital expenditure and anticipation. Subramanian cited Uber as a cautionary example, noting that the company blew through its entire annual IT budget in the first quarter without becoming four times more profitable. Instead, Uber reported an okay quarter with okay earnings growth. He highlighted a supply and demand picture where frontier model tokens are scarce and expensive, while large language model interactions are becoming cheaper. The challenge for companies is integrating these tools to increase productivity.
Harrow stated that capital expenditure remains the primary driver of valuation in the market. She suggested that investors should favor the infrastructure trade today, including chips, memory, and power. Harrow predicted a pivot point in one year when more compute becomes available, allowing companies to use tools to become more productive and profitable with less of their own capital investment. She affirmed that AI will be additive to the economy but requires more compute to come online.
Duff addressed concerns about overbuilding infrastructure, stating he does not believe it is happening based on the current demand structure. He argued that the situation remains stable as long as capital expenditure is spent and positive downstream results continue to appear, as they have throughout the year.
The discussion concluded with a note of caution regarding return on investment. While some executives report that access to more AI tools would have increased their profitability, others remain uncertain about finding clear ROI. The tension between massive infrastructure spending and the tangible productivity gains it produces remains the key variable for future market performance.