Premature generalization of features into products

One of the typical fallacies of my Engineer’s Mind.

I wanted to create a custom specialized support chat tuned for my use cases with ScreenshotOne. But then I thought, why not make it a standalone product? The scope of potential work grows tenfold and I can’t afford to build it anymore. It is a premature generalization, dumb!

I procrastinated on many features because I wanted to generalize them too often and make a separate product out of them later.

But:

1. With AI, it matters less. I can generate a ton of code, and then remove it. Tune it for my use, throw it away, or generalize later. I am not afraid to experiment. The sunk cost fallacy is not relevant anymore.
2. And in the age of AI, specialized software and taste might stand out compared to an average amorphous generic slop.
3. I can always extract and generalize later, especially once I’ve solved the problem and got some feedback on it.

Now, the real fun begins.