The article discusses the author's observation that many companies are investing heavily in AI projects, but these projects are often failing to deliver expected results. The author believes that this is due to a combination of factors, including the novelty of AI methods, the difficulty of running software projects effectively, and the tendency for companies to overhype AI's capabilities. The author also notes that expressing skepticism about AI's potential can be risky, as it may be seen as a challenge to the company's strategy. As a result, many people are reluctant to speak out against AI initiatives, even if they are not convinced of their value.