Layoffs Due to AI Are Costing Firms More As AI Bills Soar

NVIDIA executive Bryan Catanzaro says that within his team at least, AI systems cost far more than the employees they are supposed to help or replace. He says the main driver is not salaries, but the high price of compute, including GPUs, cloud time, and energy. For his own team, he says the cost of compute is “far beyond” the cost of staff, which means AI use at scale can quickly become very expensive, as reported by Tom's Hardware.

Reports highlight that this problem goes beyond Nvidia. Full automation with AI only makes economic sense in a minority of jobs that rely on tasks like image analysis, with humans still cheaper in most cases.

Many companies are finding higher costs due to increased cloud bills and subscription fees for large models that charge based on usage, like per token or per query. In some cases, small teams have seen ...