Bryan Catanzaro (Nvidia's vice president of applied deep learning) stated that the cost of AI compute has surpassed that of human employees, making AI more expensive than traditional labor. A recent MIT study supports this claim, revealing that AI automation is only financially viable for about 23% of jobs, particularly those involving visual tasks. In the remaining 77%, human workers remain the more cost-effective option. Despite the high costs and unclear productivity gains, big tech companies have committed approximately $740 billion to AI-related expenses this year, marking a 69% increase from 2025. This surge in investment comes amid rising tech layoffs, with over 92,000 workers losing their jobs in 2023 alone, a pace that exceeds last year's total layoffs. Experts suggest that while AI is currently more costly than human labor, this could change as the technology becomes more reliable and infrastructure improves, potentially leading to a future where AI is both cheaper and more predictable at scale.