In a surprising turn of events, a recent study by Aberdeen has found that businesses are no longer prioritizing computing power over efficiency when it comes to generative AI (gen AI). This shift in priorities marks a significant departure from the initial hype surrounding gen AI, which led many organizations to sacrifice power conservation and sustainability in the race to deploy powerful computing infrastructures.

The study, which surveyed businesses on their use and plans for AI, found that over 90% of organizations are using AI in some form, with 25% making dedicated strategic investments in building AI. However, instead of relying on power-hungry GPUs, businesses are opting for small custom language models that can be built on a single engineer's laptop, minimizing power consumption.

The trend towards efficiency is further supported by the fact that increased storage capacity and hybrid cloud capabilities have become the top technologies purchased to support AI initiatives. This shift in priorities is a welcome change for the tech and startup community, as it indicates a growing recognition of the importance of sustainability and power conservation in AI development.