Rising debt loads and capital expenditures exceeding $1 trillion globally in 2026 for AI infrastructure, combined with falling inference prices from Chinese open-weight models like Qwen and GLM, have fueled trader caution around an AI bubble burst. Prominent voices including Ray Dalio and Michael Burry highlight classic bubble signals such as concentrated valuations, credit stress, and expectations that hyperscaler returns may disappoint relative to spending. Recent OpenAI revenue revisions and Nasdaq volatility tied to these reports have reinforced near-term skepticism, while strong contracted backlogs and adoption metrics provide counterbalance. Key near-term catalysts include hyperscaler earnings, potential Federal Reserve rate moves, and further open-source model releases that could accelerate price erosion.