Misalignment of Goals Among Stakeholders
Emerging technologies like AI are tempting for investors, but research by the RAND Corporation highlights a concerning trend: over 80% of AI projects fail. A significant factor is the misalignment of goals among key stakeholders. Leadership often envisions AI's capabilities based on unrealistic expectations, sometimes influenced by Hollywood's portrayal of AI. This often leads to a disconnect between business leaders and the technical teams, resulting in projects lacking the necessary resources and time for success.
Engineers and the Shiny Object Syndrome
While management's unrealistic expectations contribute to failures, the engineers and data scientists are not without fault. They sometimes become enamored with the latest AI developments, a tendency known as the "shiny object syndrome." This means they might implement new technologies because they are novel, not necessarily because they solve the current problems in their projects. Staying updated with tech trends is essential, but it is crucial to focus on technologies that add real value.
Other Contributing Factors
Additional factors contributing to the high failure rate include inadequately prepared data sets and insufficient infrastructure. Moreover, the incompatibility of AI solutions with the problem at hand remains a persistent issue. The academic sector also faces challenges as many projects emphasize publishing research over practical applications, which can lead to similar failures.
Industry-Wide Implications
These findings are not just applicable to private companies. Public and academic entities also suffer from the same pitfalls. For instance, in China, Baidu's CEO Robin Li Yanhong pointed out the excessive number of large language models developed with little practical application. Although China outpaces the U.S. in AI patents, meaningful contributions are limited, as evidenced by the Chinese Academy of Sciences being the only Chinese entity in the top 20 for citations between 2010 and 2023.
Learning from Failures
The drive to outpace competitors in the AI landscape is causing many firms to rush into projects without fully understanding their potential pitfalls. For companies and investors, it is crucial to scrutinize past failures to avoid similar mistakes. If AI projects continue to underdeliver on their promises, it may lead to a significant downturn in the industry, similar to a "trillion-dollar bubble" burst.
