Tech companies are grappling with the high costs of artificial intelligence, which are now exceeding the expenses of the human labor they were intended to replace. Uber's CTO revealed that the company exhausted its 2026 AI coding budget in just four months, with 84% of engineers using AI tools like Claude Code. Despite high adoption, the value of AI-generated code remains unclear, as noted by Uber's COO Andrew Macdonald. This trend is not isolated to Uber. Microsoft, after investing $13 billion in OpenAI, halted the use of an AI coding assistant due to unsustainable costs. An unnamed company reportedly incurred a $500 million AI bill in a single month. Nvidia's VP of applied deep learning, Bryan Catanzaro, acknowledged that compute costs now surpass employee expenses, with the company planning a $2 billion annual token budget for its engineers. The financial strain is evident as tech firms lay off over 115,000 workers in 2026 to redirect resources towards AI, despite studies indicating that AI automation is only economically viable for 23% of roles. As AI spending continues to rise, the industry faces a critical question: can AI investments justify their costs before financial resources are depleted?