That sounds completely unrealistic and utopian to me. Before I even consider how to deploy a product, I want to experiment locally to find out if I even have a

That sounds completely unrealistic and utopian to me. Before I even consider how to deploy a product, I want to experiment locally to find out if I even have a potential product. Most of my experiments never amount to being anything more than an experiment. If I had to set up a CI pipeline before writing any code, most of them wouldn’t even amount to that! Maybe, if I were executing on someone else’s vision, I could justify taking time up front to set up CI. But I still think it’s a little dogmatic to state that this has to be the first order of business, always.