Uber's COO says it's getting harder to justify the money spent on AI tokenmaxxing

In a recent interview, Andrew Macdonald, Uber's Chief Operating Officer, expressed concerns regarding the company's spending on artificial intelligence. He highlighted that justifying the costs associated with AI has become increasingly challenging for the business. Macdonald referred to a statement made by Uber's Chief Technology Officer, Praveen Neppalli Naga, who previously noted that the company had already exceeded its budget for AI development, specifically the Claude Code budget for 2026. This revelation prompted what Macdonald described as a 'head-exploding moment,' igniting discussions around AI token consumption and the consequent trade-offs, particularly concerning staffing levels. He emphasized that conversations with senior engineering leaders revealed a disconnect between the increased usage of AI tokens and the enhancement of consumer features. "That link is not there yet, right?" Macdonald remarked. He added that, while there may be more features being developed, establishing a clear relationship between AI token usage and the delivery of substantial consumer benefits remains elusive. The COO's comments come on the heels of an earnings call where Uber's CEO, Dara Khosrowshahi, announced a slowdown in hiring to manage the company's AI investments effectively. Macdonald noted that while AI may seem cost-free from a user's perspective, the financial burden ultimately falls on the company. In the broader tech landscape, companies are reassessing their approach to AI utilization. For example, Duolingo recently reversed its decision to incorporate AI usage into employee performance evaluations after feedback indicated that employees felt pressured to use AI without clear objectives. Duolingo's CEO, Luis von Ahn, articulated that the focus should be on tangible outcomes rather than merely increasing AI engagement for its own sake.