The author argues that if companies merely provide AI tools and encourage experimentation without first defining the business problem, ownership, and measurable objectives, experiments are unlikely to produce returns and may weigh on profits. The article cites the results of MIT’s NANDA initiative’s review of more than 300 enterprise AI projects, finding that 95% of generative AI pilots generated no measurable return on an estimated $30 billion to $40 billion in corporate investment. Gartner, meanwhile, forecasts that more than 40% of agentic AI projects will be canceled by the end of 2027.
Source:The AI Journal · Read Original · 前沿热点
Leave a Reply