• Welcome to AppraisersForum.com, the premier online  community for the discussion of real estate appraisal. Register a free account to be able to post and unlock additional forums and features.

20 Minute AI Appraisals Are Coming

I totally disagree with you. I will believe it when I see it.
Im sure its very good at spitting out a result that looks like. Market conditions change, real estate is local, and data is not standardized (agents putting in basements in square footage, enclosed porches as GLA, etc). Also data from public records isn't that great especially for cape cods and split levels. I'm sure there are many other things, especially market specific things I don't even know about from other markets.
 
You may have missed the part about AI not handling external well yet. It will, though. Teaching it to perform paired sales analysis is child's play. Teaching it to recognize differences in 'deep water' access, rocky shores, 'open water', etc. is going to be a challenge.
I don't know how well it works, but I can say I saw an AVM for a refi of my parents house and it was awful.

2 story colonial subject in a water privileged community,

It skipped over very similar sales in the subject's community to take a split level that is less than half the GLA across a major road (different market) that isn't in a water privileged community.

If an appraiser did that it would be malfeasance without a good explanation.

If I had to guess it was weighting the date of sale way too highly, that sale settled within 1 month whereas the very similar comps in the community sold 3-4 months ago.
 
Last edited:
Very good point....
For that 1 property....
20 human appraisers can bid for the assignment....
and still not properly understand nor adjust for water depth, amount of shoreline, what size a dock if permitted, or...........etc. etc.
 
I don't know how well it works, but I can say I saw an AVM for a refi of my parents house and it was awful.

2 story colonial subject in a water privileged community,

It skipped over very similar sales in the subject's community to take a split level that is less than half the GLA across a major road (different market) that isn't in a water privileged community.

If an appraiser did that it would be malfeasance without a good explanation.
I've no doubt that's the case.
 
That's what Retrieval-Augmented Generation (RAG) and Cache-Augmented Generation (CAG) is for. You can feed AI the same information available to any appraiser and teach it what to do with it.
RAG still depends on quality data and even then, is highly variable in terms of it's "analysis" and output. I paused building a side project because the tech was just not there (especially PDF parsing - in particular, tabular data within PDFs).
 
If AI truly can "do" an appraisal in the future, it is not an appraisal if a human does not develop the opinion. It is some alternate form of valuation.
Lenders are not going to care that you don't think that an AI generated appraisal should be called an appraisal.....if and when lenders determine that AI generated appraisals are reliable and are cheaper/more efficeint than human appraisers, then lenders are going to use AI, it is as simple as that. I don't think human appraisers will be totally eliminated....they will still be needed for atypical properties and properties in rural areas with little available sales data (rural mortgages are <10% of total mortgae volume), however, AI is very likely to replace human appraisers at some point in the future when it comes to the valuation of most condo units, and tract homes. I wish this was not the way it is going to be, but hopes and wishes are of no value.
 
Find a Real Estate Appraiser - Enter Zip Code

Copyright © 2000-, AppraisersForum.com, All Rights Reserved
AppraisersForum.com is proudly hosted by the folks at
AppraiserSites.com
Back
Top