LLMs are the VisiCalc of Knowledge Work
Declining marginal cost = increasing volume of slop
In 2024 I wrote about how the invention of VisiCalc was the “ChatGPT moment” for data analytics and financial analysis. A new tool made it so you didn’t need an army of bookkeepers and paper spreadsheets to do data-driven, deterministic work. As the marginal cost of producing analytics went to (near) zero, we naturally saw an explosion of accountants and financial analysts.
As VisiCalc (and eventually Excel) proliferated, what remained as unfair advantages were relationships, insights, information (collectively, “Intangible Assets”). If you had better information on a business’ growth prospects, your Excel model would prove more accurate. If you had unique relationship, your purchase price in the model might be advantaged.
The same thing is happening now with AI and knowledge work. If we trust the analogy, then we know the volume of knowledge work is about to explode (I recognize the irony as I write a weekly blog, full of words). As a result, insights, information, and relationships are more valuable than ever.
In this new paradigm, it’s a company or investor’s Intangible Assets that will create an unfair advantage—unique information that feeds the tool, the way one uses the tool to move faster, and the relationships and reputation one builds that AI cannot automate away.