To predict AI's future impact on the broader economy, observe its current capabilities in software development. AI models are consistently about a year ahead in coding ability compared to other domains, providing a reliable preview of the automation coming to other knowledge-work sectors.

Contrary to common belief, AI's initial impact is on white-collar roles like analysts and writers. The real bottleneck in the AI revolution is a shortage of skilled trades. Nvidia's CEO stated the biggest hurdle for data center construction is finding enough plumbers.

Financial analysts are modeling AI's economic impact using a flawed, zero-sum perspective, similar to early estimates for PCs and the cloud. They're missing that AI will create entirely new business models and drive a 1000x increase in resource consumption, making the total opportunity orders of magnitude larger.

Massive, long-term investment in AI data centers assumes current power models will persist. Future AI efficiency breakthroughs could render many of these facilities obsolete or underutilized, similar to the overbuilt fiber optic networks of the dot-com era.

AI tools make software development drastically cheaper. Rather than replacing engineers, this efficiency will likely trigger the Jevons paradox: the unlocked demand for new, more powerful software will skyrocket, increasing the overall need for people who can direct these new capabilities.

The future of compute demand is a tale of two opposing forces. Enterprises will use AI to compress redundant data and streamline operations, reducing compute costs. Consumers, however, will demand generative AI for entertainment and personalization (e.g., 'Star Wars with my face'), creating massive new compute needs.

The narrative of insatiable AI compute demand is partially a bubble. It's fueled by inefficient early models ("token maxing") and a culture where tech executives brag about their AI spending as a status symbol, a behavior not seen with traditional cloud costs. This suggests demand could normalize.

The AI productivity boom is confined to tech because developers have fewer adoption hurdles. Coding is a text-only medium with self-contained context in a codebase. In contrast, roles like marketing or law require complex data setup and workflow re-engineering, slowing down the productivity gains seen in macro-economic data.

The builders of AI may have a skewed perspective on its real-world impact. They often extrapolate from their tech-centric experiences and fail to grasp how technology diffuses in the broader economy. Their predictions about societal consequences, such as mass job displacement, should therefore be viewed with healthy skepticism.