Recently, an Nvidia executive highlighted that the cost of compute has now surpassed employee salaries. Last week, Mercor's CEO stated that the startup spends more on tokens for internal agents than on its employee headcount. As enterprises exhaust their token budgets, a pivotal question arises: Are companies actually spending more on AI than on human resources? According to fresh research from the Ramp AI Index, the answer is not yet. The top 1% of firms, termed as 'AI-pilled' by Ramp, spend $7,500 per employee per month. Whether this is considered high or low is subjective, but it is certainly less than the average software engineer's salary of about $16,000 per month. This statistic pertains only to power users. The top 10% spend approximately $611 monthly per employee, while the median expenditure is just $11.38, roughly equivalent to the cost of a seat on an enterprise plan. Despite pressures, AI spending continues to rise. Among AI-pilled firms, spending per employee grew by 14.1% last month. It remains uncertain whether this trend will persist. The top 1% of firms tend to mix and match, opting for multiple frontier models and platforms that provide access to cheaper open-source models.
Blogger's Review: The rapid advancement of AI technology is leading to significant increases in corporate investment in AI, though it still lags behind human costs. As the technology matures, whether AI spending will eventually surpass that of human resources is a trend worth monitoring. Companies must consider how to effectively leverage AI to enhance productivity while optimizing resource allocation.