I had previously discussed the overhyped LLM market. However, it must also be recognized that with significant advancements in the field of AI and LLMs, there could not be a more opportune time to examine Jevons Paradox, which explains how technological improvements that increase efficiency result in greater, rather than reduced, overall resource consumption.
Seemingly, traditional economic wisdom purports to suggest that when the supply is more, the demand is usually less. But Jevons Paradox is an exception to this traditional wisdom. It says that when the supply increases, the costs lower, and consequently, this lowering of costs boosts the demand for consumption. This paradox is not new and has also been used to describe other sectors like energy and infrastructure.
What we need to understand is that an increase in technological efficiency affects not only the aspect of supply but also alleviates the overall quality of the resources in question. In my humble opinion, it is this qualitative aspect that plays a critical role in boosting the consequent demand side of the concerned resource.
Usually, technological improvements are incremental in nature, but if such incremental pace is sustained over longer periods of time, then it may have a positive lasting impact on the supply side of the resources.
The point is that with DeepSeek, Grok, Manus AI, etc., in the market, this race for efficiency is only going to increase. Better and efficient AI algorithms are fuelling Jevons Paradox in a massive manner, more so than the rate at which the computing hardware is witnessing progress.
Though both Nvidia and AMD are leaving no stones unturned to take out as much juice as possible for the current state of hardware, yet with more efficient AI algorithms, the efficiency improvements are much more palpable. Today, it is possible for any consumer-grade laptop or desktop to locally install a distilled version of popular LLMs. This is a huge stride, and it took very less time to achieve the same.
What does this tell us about the future? We are moving towards AGI, artificial general intelligence. Human-like robots, fully automated businesses, and systems with zero human interventions are just around the corner.
As Jensen Huang, Head of Nvidia, says, this is the age of agentic AIs. At first, it sounded like a market gimmick to many, but what Jensen was implying was profound. All our jobs are going to be replaced by AI agents. This does not mean that there will be no jobs in the future; rather, it would mean that the nature of job and work is simply changing. And like Moore’s Law, how long will Jevons Paradox sustain? Nobody knows. So far, it is not showing any signs of slowing down. Truly exciting times to live in, let’s make the best out of it, hopefully.