PART III: Routes to defensibility for your AI Startups

One graph and three variables to rule them all

I have had many conversations with founders or VCs about how to define and assess the defensibility of an AI startup since I published my initial posts. In case you missed them, see Part I and Part II.

Having another one today, I just realized that we can sum it up in a simple graph with three variables.

Let’s define:

• The Minimum Acceptance Level (or MAP) as the minimum level of data, time or funding required for users to adopt the software (more about the MAP here).
• DMAP as the the amount of data or time required to get to the MAP
• μ as the value of the derivative function of the learning curve at DMAP

Results: The larger DMAP and the smaller is μ, the more defensible is the company.

A bit of maths but I find it cool that three variables are enough to think about the concept.

Thanks @Gvinzelles, it was an interesting conversation!