AMA Epistemology Answer: Limits of Testing

Partially back from a two month hiatus - apologies this took forever but it never left my backlog!
From the original moment: AMA about Epistemology - the practice of testing

Simon asked: "Can you share some examples where epistemology highlights the limits of what testing can actually prove?"

I had to think about this one! I hope I am answering the right question - "testing" is a nebulous term. 

Epistemology is a science (hence, -ology), and therefore follows the scientific method: Observation, Hypothesis, Experiment

You see a button, you get an idea of what the button does based on context, you click the button. Repeat: you observe what clicking the button did, consider why it did that, and then try again possibly with different inputs.

The key to answering this is within that process. In order for us to test something, we must be able to use the scientific method. It must be a real, tangible thing. You need a running product. An idea is just a belief - something that may be true eventually. Requirements documents, acceptance criteria, market research, wireframes - all illustrations of an idea which may well prove a fever dream. You need a system that makes that belief falsifiable. Until you have a system that can show you the truth of what you built, all you have is ideas. And ideas cant be tested.