A recent article from Inc., “The Coasean singularity: Why AI is ending the org chart as we know it,” by Howard Yu, posted on https://www.inc.com/howard-yu/the-coasean-singularity-why-ai-is-ending-the-org-chart-as-we-know-it/91337326, May 6, 2026, cites a paper from 1931 that laid out a key principle relating the size of the firm to the internalized cost of coordinating major decisions. The current author’s assertion is that coordination costs have been dropping over several decades and are now accelerating due to AI. Yu gives an example of a large manufacturing entity fragmenting its production into semi-autonomous, product-specific units, and achieving greater economic efficiencies as well as increased worker satisfaction in the process.
In what would seem to be a parallel event, Nvidia has a paper promoting their concept of how firms in different types of industries can now build their own internalized “AI factories” using NVIDIA platforms, which Nvidia claims can be firms’ most efficient way for “unlocking the full potential of generative AI and reasoning AI. . .” while also giving firms greater security and flexibility in the use of their data and analytical processes. (“Unlocking Business Value With Generative AI and AI Reasoning,” https://www.nvidia.com/en-us/lp/ai/ai-factories-in-action-ebook/, accessed May 7, 2026).
We have also seen in real time recent mass layoffs by major tech firms, and we can all speculate to what extent these “liberated” workers may reemerge as individual enterprises of some type.
If firm fragmentation is in fact a real thing driven by some of these situations, what are the implications for economic development? Another line of speculation would be that places that cannot realistically compete for very large enterprises might prove to be quite efficient as locations for smaller manufacturing, software development, or other enterprises. To cite just a couple of arguments, costs of production could be lower and workers might be in a better position to experience a higher quality of life.