Investor warnings from figures like Ray Dalio and Michael Burry have intensified scrutiny of AI infrastructure spending, with hyperscalers issuing hundreds of billions in debt amid rising bond yields and questions over sustained returns. Large language model pricing has fallen sharply while data center costs remain elevated, prompting analysts to flag 2027 as a potential inflection point for corrections. Recent model releases, including Google's Gemini 4 Argon and OpenAI's agent advancements, demonstrate ongoing capability gains, yet enterprise ROI timelines and power constraints add uncertainty. Traders weigh these dynamics against historical bubble precedents, with skin-in-the-game odds reflecting debate over whether transformative adoption will outpace financing pressures before major pullbacks materialize.