AI Won’t Make Money, Like Full-Text Search Didn’t in the Late 1990s
The Market For Knowledge Is Miniscule
Knowledge is power. In the late 1990s I thought I was going to make a killing developing “full-text” search systems.
Before 1993, investment research on Wall Street meant calling a corporation and having them mail you a prospectus or annual report. Then the S.E.C. developed EDGAR (Electronic Data Gathering, Analysis, and Retrieval). The goal was to make its auditing work easier. Since these are public documents the S.E.C. made the electronic version public for a small fee (relatively speaking). With a grant from the National Science Foundation, the Stern School of Business, at N.Y.U., created an online system for public access. When the grant ran out everyone clamored for the S.E.C. to continue public access, which they did.
In 1993, I started a company, PresenText. My competition was Disclosure, which scanned financial documents and archived them on CD-ROMs. The service was very expensive.
To make a long story short, I didn’t succeed because my assumption was wrong. Financial analysts were NOT spending all their time pouring over financial data. The big brokerages made money doing public offerings. To this day, financial analysis remains mostly window-dressing. Wall Street makes its money as…