Uber has not found clear evidence that increased use of artificial intelligence improves the customer experience, Andrew Macdonald, the company’s president and chief operating officer, said on the Rapid Response podcast.
Macdonald’s remarks followed disclosure that Uber had already spent its annual budget for AI by the end of April. He warned that if the company cannot demonstrate a direct link between AI spending and the appearance of useful features for users, it will become harder to justify larger AI expenditures. "This connection is not very visible yet," he said.
In early May, Uber CEO Dara Khosrowshahi announced that the company is funding AI investments in part by hiring fewer employees. Referencing that approach, Macdonald said: if they cannot clearly show how many useful features are delivered to users as a result, it will be increasingly difficult to justify the shift.
The issue is not unique to Uber. Cybernews notes that several other large firms have recently questioned whether AI delivers as much value as the money it consumes. Target, for example, said it is reevaluating AI deployments and plans to implement or expand AI only where it provides genuine assistance. Starbucks withdrew an AI-based inventory system introduced nine months ago after it repeatedly miscounted and misidentified items.
Macdonald also quoted Brian Chesky, CEO of Airbnb, who observed that while such developments can be expensive, it is not obvious that consumers are better off if platforms switch to AI-driven interfaces.
The discussion highlights a practical challenge for companies: they must produce measurable, direct evidence that AI investments improve user experience, otherwise it will be harder to sustain or increase spending on AI.
Key facts
- Uber had spent its annual AI budget by the end of April.
- Andrew Macdonald: “This connection is not very visible yet.”
- Dara Khosrowshahi said in early May that AI investments are being funded by hiring fewer people.
- Other major companies (Target, Starbucks, Airbnb) have also revisited or rolled back AI solutions after they failed to meet expectations.
