Uber President: AI Spending Becoming Harder to Justify
The President of Uber stated that the company currently sees a lack of clear correlation between AI usage and productivity improvements, making it increasingly difficult to justify AI spending.
Despite a high adoption rate of AI tools internally, no significant improvements in personnel efficiency or business outcomes have been observed.
This statement reflects a realistic assessment by some large enterprises regarding the return on investment in AI, shifting from early enthusiasm to more cautious ROI validation.
Source: Public Information
ABAB AI Insight
Uber has previously been proactive in deploying AI in areas such as autonomous driving, pricing algorithms, and customer service. The President's public questioning of AI productivity returns continues its cautious evaluation of new technology implementation, similar to earlier reflections on the costs after long-term investments in autonomous driving.
On the capital front, Uber is reassessing its AI budget allocation, shifting resources from general AI tools to specific applications that can directly enhance driver utilization, order matching efficiency, and user retention. The motivation is to control costs through strict ROI validation and avoid high computational expenses becoming a long-term burden.
Similar to Karri Saarinen's previous concerns about AI productivity not translating into revenue growth, and cases like Microsoft reducing access to external AI coding tools, Uber is currently in a phase of adjusting from widespread adoption of AI across the company to targeted investments in specific scenarios.
Essentially, this is a concentration of capital: AI productivity improvements have yet to effectively penetrate core business processes, as existing organizational structures, workflows, and incentive mechanisms continue to operate in traditional ways, resulting in much AI usage remaining at the experimental level. Capital is shifting from broad spending to vertical scenarios that can clearly quantify returns.
ABAB News · Cognitive Law
Without productivity realization, AI spending is just an expensive experimental tax.
Adoption is easy, but truly converting it into profit is extremely difficult.
Top companies are not those that use AI the most, but those that are best at stopping ineffective AI.