AI Costs Rise as Companies Question Returns on Investment
Companies that rushed to adopt artificial intelligence are now facing a reckoning: the technology is becoming expensive, and the promised productivity gains haven't materialized.
Uber's chief operating officer recently noted that heavy spending on AI has produced no noticeable increase in productivity at the company. The observation reflects a broader shift in how businesses view their AI investments after years of aggressive spending.
The subsidy era is ending
For the past two years, major AI companies used a familiar Silicon Valley tactic: charge low prices to attract customers. Investors footed the bill, allowing firms to offer AI services at a fraction of their true cost.
Kevin Simback, a leader at start-up incubator Delphi Labs, calls this the era of "subsidised intelligence." That period is ending.
OpenAI and Anthropic, the two dominant players in the market, are preparing to go public later this year. Both companies need to demonstrate they can turn a profit, which means prices are rising across the board.
AI agents drive costs higher
AI agents are the primary reason for the price increases. Unlike chatbots that simply answer questions, agents perform actual work-booking appointments, writing code, managing files.
A single task can spin up dozens of agents working simultaneously, each generating charges. The cost structure makes agents far more expensive to run than earlier AI tools.
As pricing pressures mount, companies will need to demonstrate clear financial returns from their AI spending. The era of cheap experimentation is over.
Your membership also unlocks:
