Have you seen the latest report released by MIT?
The shock of the report lies in its empirical data that shatters the romantic fantasies of AI investment.
Through interviews with 150 business leaders and analysis of 300 cases, the report reveals three disruptive conclusions:
95% of generative AI projects yield no returns: the success rate for enterprises building their own systems is only 33%, while the success rate for purchasing third-party solutions is as high as 67%. This directly challenges the business models of companies like NVIDIA and Palantir that build their own AI ecosystems, leading to stock price declines of 3.5% and 9.4%, respectively.
Structural misalignment of ROI: over 50% of AI budgets are allocated to sales tools, while the actual highest returns come from backend process automation (such as reducing BPO costs). This explains why data processing platforms like Databricks outperform marketing AI tokens.
Hidden costs of organizational change: projects led by frontline managers have a success rate 2.3 times that of central laboratories. This suggests that platforms supporting decentralized AI deployments, like Coinbase, may have more long-term value.
MIT "The Generative AI Gap: The State of Commercial AI in 2025"
Report source: https://nanda.media.mit.edu/
Registration required; if you want to see it, you can contact me
