What was your take on where they're going with that product?
I get a lot of people from overseas clicking on my LinkedIn, some with Ph. D.s in Finance, and others in IT. I imagine some programmers in India, Eastern Europe, or other locations are being pressured to write programs for MARS and RCA by someone, or they may be doing this out of self-motivation.
I don't know what Jeff is up to, I have virtually nothing to do with him nowadays - except I bump into him once in a while when he is giving one of his simple presentations at some AI conference. However, he has a reputation for not following through on his "avant-garde" ideas. Mostly, it seems, he keeps modifying his very old (though very high-quality) Pascal-based software from eons ago. He does experiments, but often throws everything away for one reason or another. Understand, I say this mostly from my experience working for him for one year back in 2007-8, Glassdoor comments from others who have worked for him, and occasional posts we sent back and forth maybe 10 years ago for a short period when he was working on AI for photos. Jeff certainly tries to keep himself in the loop and he has a ton of contacts. But, having contacts in Appraisal - only gets you so far in developing new technology; in fact, any kind of contacts in the "Appraisal World" will most likely constrain the developer. And so, from my point of view, he works in a world where he feels he has to follow bad advice - something I refuse to do. But then again, he has, up until now, always had enough success to keep his business going. If I had to guess, though, -- and honestly, I don't know for sure -- I can't imagine he does that well from his appraisal business. I always thought he had an outside source of income - perhaps from his other investments. But I can only guess. I don't know or have that much interest in his doings, at least outside of what he is trying to do in AI Appraisal, -- and I think that is really going to depend on his resources whatever they are; but I would guess he is running a tight ship and depends on convincing certain others with deep pockets, that he is on the right track.
The main problem anyone is going to run into now:
1. Almost no appraisers can handle MARS or anything equivalent right now. And on this forum and elsewhere, they will state they want a "one-button push system." However, anyone who works with advanced appraisal tools will tell you that in this field "one-button push systems" don't work very well. You need a talented and experienced programmer/appraiser to oversee the entire project. One who knows right from wrong. I would expect the vast majority of entrepreneurs who make attempts in this direction, will eventually give up. It is hardly doable for reasons outside their own competence. - Too many loose ends.
I would build a system for a 140 IQ appraiser, for the future. It wouldn't be a one-button system - it would be a very sophisticated tool set, probably with very well-written R and extensive libraries written in C++ that could be called from R. Lots of graphics. LOTS of documentation and examples. But you need a developer who has an in-depth understanding of appraisal.
HOWEVER, such a system could also be written for other areas in finance, and if I had to guess, we will see these other systems sooner than appraisal systems.
For myself, I spend my time developing infrastructure. I am basing everything on the Apple/MacOS/PostgreSQL ecosystem to provide a highly integrated "Kit" for a highly advanced appraiser. So, for example, I am inclined to use the top MacOS tools such as Alfred 5, which make it easy for an advanced analyst to write their own workflows quickly and easily --> to be able to adapt to changing SOWs.
2. A host of associated problems, the system inertia of the appraisal profession as a whole. ... More some other time.