Appraisal doesn't pay much anyway, especially for residential specialist. The incoming AMC fees are low and the outgoing fees for everything else add up to push you into debt.
Anyway, in the name of transparency, we also need open-source appraisal software. So, my conclusion is we need more open software for appraisal. And then the question is how software developers can make a living off writing high-quality open-source software. AI makes it easier of course. For $200/month, a good software developer essentially has a team of assistants to do the run-of-the-mill coding, such as writing unit tests. My current "earthUI" now has about 1,000 unit tests - and can keep them updated as I make changes. It supports different date and formatting standards for about 30 countries, including Japan and South Korea. I could make it support the different languages as well, only right now I have more important things. It also takes care of complex Excel formatting for Sales Grids that can contain an unlimited number of sales comparables - all which will have the same exact adjusted sales price. And so on and so on.
Also working on glmnetUI, that supports a linear model WITH INTERACTIONS. I bet you have never used that! There are some market areas where you can create some very good linear models, especially if you add interactions, such as living_area vs total_baths.
And working on mgcvUI, which can (optionally) feed earthUI output into mgcv() to create smoothed/curvilinear models like earthUI.
And then the older ValEngr app is much more complex and larger.
And more apps on the drawing board, plus my own publication system. Maybe someday I will make money off all of this. In the meantime, I enjoy the work.
Note for one thing, based on my years working in Germany as an International Accountant (Westinghouse, New England Nuclear, Nippon Electric Corporation) - I intend to make everything international and supply it to the whole world, which should open up the market.