Top business leaders are expressing growing uncertainty about the financial viability of artificial intelligence investments, with Uber's operations chief Andrew Macdonald now questioning the trade-offs of AI investments within the company. In a recent Rapid Response interview, Macdonald said it was becoming harder to justify AI costs, particularly as he's not seeing proportional productivity gains from increasing AI costs. This follows Uber's Chief Technology Officer Praveen Neppalli Naga's viral comments in April to The Information that the company had already burned through its Claude Code budget for the whole of 2026. Macdonald described this as a 'head-exploding moment' that sparked discussions about AI token consumption within the company and the trade-offs it creates, including impacts on head count.
Economic analysis suggests AI's potential for productivity improvement may be more limited than commonly assumed. According to MIT Professor Daron Acemoglu's research titled 'The Simple Macroeconomics of AI', AI can cost-effectively automate only about one-quarter of tasks it can technically handle, translating to roughly 5% of all tasks overall. This translates to a total productivity gain of 0.5% in the US and will contribute just 0.9% of US GDP in 10 years. Macdonald specifically noted that higher token usage did not translate into a proportional increase in useful consumer features, stating 'That link is not there yet' and adding that 'it's very hard to draw a line between one of those stats and, 'Okay, now we're actually producing 25% more useful consumer features.'
The scale of AI investment commitments has reached unprecedented levels, with significant implications for future returns. As reported by Where's Your Ed At newsletter, hyperscalers have invested over $800 billion in the last three years, with plans to add another $700 billion in 2026 and another $1 trillion in 2027. This massive capital deployment means companies need to generate at least $3 trillion in AI-specific revenue just to break even, with $6 trillion or more required for AI to deliver meaningful returns beyond breakeven. Macdonald added that AI can seem free to users creating interesting use cases without paying for it, but 'ultimately, the company foots the bill.'
A separate EY survey identified significant execution gaps that are limiting AI's financial impact. The survey found that companies are missing out on about 40% of potential productivity gains due to weak strategy and execution. According to PwC, these challenges have resulted in declined confidence among CEOs about revenue growth, with only 30% saying they are very or extremely confident about revenue growth over the next 12 months, down from 38% in last year's report and significantly lower than the peak of 56% recorded in 2022. The lack of clear strategy may explain why companies don't know their quarterly AI costs, leading to experimental and open-ended spending that can inflate reported revenues for firms like Anthropic.
