“What did we get wrong?”

It’s such a useful question, but one many businesses skirt.

Union Square Ventures is one of the exceptions. 

The venture capital firm co-founded by Fred Wilson is taking advantage of AI to ask that and other questions of their own archives. 

Matt Cynamon, head of library sciences at Union Square, recently shared on Dan Shipper’s podcast that the firm has created a database of 15,000 articles from its blog.

AI lets partners: 

–Ask and answer questions based on previous blog posts

–Track and analyze portfolio investments 

–Review notes taken at meetings 

According to a post on X, Matt said the firm analyzed posts between 2008 to 2012 to find examples where they missed an opportunity.

“In a post from October 2, 2012 titled “Researching Online Education”, Christina Cacioppo shared some of USV’s thoughts on the online education market. One key point was: “We’re skeptical a business model that charges for content will work at scale and in the long run.” 

This skepticism about paid content models for online education turned out to be incorrect. In the years following 2012, several successful online education platforms emerged with business models that do charge for content, such as Coursera, Udacity, and edX. These platforms have achieved significant scale and longevity while maintaining paid content as a core part of their business model.”

It’s an evocative example because it illustrates one of the most compelling use cases for AI.

Namely, to tap into reservoirs of specific corporate data or private information and ask narrow, actionable questions.

This increases the utility of the information and decreases the likelihood of hallucinations.

Previously, this kind of analysis would have required too much manual effort.

Increasingly, the only obstacle will be the willingness of a company to ask hard questions.