Bank of America said the concentration of artificial intelligence-related investments in U.S. stocks resembles the market structure before the dot-com bubble burst in 2000. Michael Hartnett, the bank’s chief investment strategist, said investors are favoring AI-linked assets including the Nasdaq 100 and the Magnificent Seven while overlooking the equal-weighted S&P 500. He compared the current boom with 1999, when technology stocks rose more than 40% in the six months before the March 2000 peak while consumer staples fell more than 30%. Hartnett called AI the biggest bubble since railroads, although he said it is too early to conclude the boom has peaked. Capital spending by hyperscale cloud companies is projected to reach 3.5% to 4% of U.S. gross domestic product next year, below the roughly 5% recorded during the 19th-century railroad construction boom. Unlike the late railroad cycle, semiconductor prices are still rising, but today’s market faces high interest rates rather than falling Treasury yields. Hartnett recommended gradually increasing bond allocations and pursuing a buy humiliation strategy focused on assets that have been heavily sold off and shunned. The 10-year Treasury yield recently reached 5.33%, its highest level since 2002, while rising yields reduce bond prices.