What AI Owes the People Who Explained Commercial Real Estate
Every answer your AI tool gives you about cap rates, lease structures, or the debt markets came from somewhere. Not from the model – from a person who explained it in public first.
In 1675, Isaac Newton put it best in a letter to Robert Hooke: “If I have seen further it is by standing on the shoulders of giants.” Everybody starts from somewhere. The view is better from up there because somebody else did the climbing.
It’s a fair description of what AI is doing right now.
Every answer has a source, even when you can’t see it
When a language model answers a question about lease structures or cap rate compression, it isn’t reasoning from first principles. It’s drawing on articles, transcripts, panels, forum threads and the accumulated writing of people who explained the industry in public – names like Peter Linneman, Spencer Levy, and Willy Walker, among others whose research and commentary shaped how the field talks about itself.
None of them set out to train a model. But their transcripts and notes became part of the material AI learned from. Ask a question, get a fluent answer in two seconds, and it feels like the machine knew it. It didn’t. Somebody taught it – usually without knowing they were doing so.
Why this matters for how you use the tools
Two things follow.
First: a model is only as good as what was published before it. Where an industry has argued in public and shown its work, the answers are strong. Where knowledge stayed in someone’s head or a deal file, the answers are sparse – which is why AI can discuss cap rate theory fluently and still tell you nothing about why occupancy slipped at your property last quarter.
Second: your own data is the part nobody else has. The giants gave the industry its general knowledge. Your portfolio, your variance explanations, and your property teams hold the specific knowledge – the reason a particular GL line moved isn’t in any transcript, it’s in what the people closest to that building actually saw happen. No amount of general training substitutes for it. The most useful thing you can do with these tools is point them at the ground truth you already own.
AI gives you the accumulated view of everyone who explained the industry in public. It can’t give you the view from inside your own buildings. That one’s still yours to supply.
Credit where it’s due
Remember the answer had authors. Read the research, listen to the panel, follow the people still doing the climbing – because the next generation of models will train on whatever they publish next.
But the next frontier isn’t just public AI models answering general questions. It’s AI built into your own platforms, trained to listen for the same thing the giants have always provided – an explanation, not just a number – except this time from the people standing closest to your buildings. Refined Data is one place we’re putting that thinking into practice.
So: who’s on your list? Which voices in commercial real estate do you actually make time for – the podcasts you finish, the research you read the week it lands? And just as important: are you capturing the voices closest to your own buildings with the same discipline?
—
References:
Peter Linneman: https://www.linkedin.com/in/peterlinneman/
Spencer Levy: https://www.linkedin.com/in/spencerglevy/
Willy Walker: https://www.linkedin.com/in/willy-walker/
