A new analysis argues that despite rapid advances in artificial intelligence, human labor remains more cost-effective than large-scale AI systems for many real-world tasks, largely due to the high infrastructure and energy demands of computing.
According to commentator Rob Enderle, the long-held belief that automation would always be cheaper than human workers is beginning to break down, as modern AI systems require significant power, cooling, and computational resources to operate effectively.
He notes that while AI can process tasks at scale, the cost of training and running large models often exceeds the expense of employing human analysts for comparable work.
In some cases, replacing a mid-level human role with AI may generate poor returns once infrastructure, error correction, and risk management are included.
The report also highlights the efficiency of human biology, arguing that people operate on relatively minimal daily energy compared to the megawatts required for high-performance AI systems handling real-time reasoning.
However, the analysis suggests this cost advantage may not last. As smaller and more efficient AI models and specialized chips emerge, the price of digital labor is expected to decline significantly over the next decade, potentially shifting the balance again in favor of automation.
For now, though, the article concludes that humans remain the more flexible and economically efficient option—particularly in roles requiring judgment, creativity, and adaptability.
What’s your view, will AI eventually become cheaper than humans across all jobs, or will human labour always retain an edge in certain fields? Share your thoughts in the comment section.
