Pod Insights The Real Estate Espresso Podcast Asking AI A Rambling Question About this episode On today's show we are examining the best way to ask a question for maximum benefit. This is a real life first hand example.--------------**Real Estate Espresso Podcast:** Spotify: [The Real Estate Espresso Podcast](https://open.spotify.com/show/3GvtwRmTq4r3es8cbw8jW0?si=c75ea506a6694ef1) iTunes: [The Real Estate Espresso Podcast](https://podcasts.apple.com/ca/podcast/the-real-estate-espresso-podcast/id1340482613) Website: [www.victorjm.com](http://www.victorjm.com) LinkedIn: [Victor Menasce](http://www.linkedin.com/in/vmenasce) YouTube: [The Real Estate Espresso Podcast](http://www.youtube.com/@victorjmenasce6734) Facebook: [www.facebook.com/realestateespresso](http://www.facebook.com/realestateespresso) Email: [podcast@victorjm.com](mailto:podcast@victorjm.com) **Y Street Capital:** Website: [www.ystreetcapital.com](http://www.ystreetcapital.com) Facebook: [www.facebook.com/YStreetCapital](https://www.facebook.com/YStreetCapital) Instagram: [@ystreetcapital](http://www.instagram.com/ystreetcapital)
Transcript 0:01 Speaker 00 Welcome to the Real Estate Espresso Podcast, your morning shot of what's new in the world of real estate investing. I'm your host Victor Manash. On today's show, we're talking about linear thinking versus transformative thinking associated with an artificial intelligence tool. I'm going to give you a firsthand experience of learning how to use AI in a transformative way. The task at hand was to review a 49-page legal agreement that was drafted by a high-priced law firm that we hired out of New York City. The document was frankly difficult to read. It was full of indirection. What I mean by that is the document was full of references. In some cases, the definition section of the agreement would say that the actual definition was contained in some other paragraph elsewhere in the document. That means having to jump
0:45 Speaker 00 from one section of the document to another, back and forth again, just to read a single sentence. It's painfully slow and exhausting to read a document that's been structured that way. Second issue is the document made no less than 40 references to outside government regulations, including specific sections of federal and state tax codes. Here too, just reading a single sentence would mean that each time you encounter an external reference, you would have to conduct a search to look up what that specific IRS code is all about. If you're searching using a conventional tool like Google, you'll be presented with a list of potential options, and then you need to crawl through all of those and choose the one that you believe is the most appropriate description of the IRS code. Is it the code itself? Is it an interpretation bulletin? Is it a white paper from an accounting firm?
1:30 Speaker 00 And once you find the best source, then you read the five to 10 pages associated with that external reference before returning to the original document. Then imagine doing that 40 times throughout the 49-page document. The process is extremely tedious. Imagine that the independent legal counsel for the other party is undertaking this work at 600 or 900 an hour. The cost to read and understand that agreement is just mind-boggling. The next best thing you could do is to ask an AI tool like ChatGPT to provide the information you're seeking. This is essentially using an AI tool like a more focused search tool. That is still linear thinking. There's definitely a productivity gain associated with this approach. You're taking something that a human would do and doing it faster.
2:15 Speaker 00 But there's a limit to how much faster. In this case, you're making 40 separate requests of the AI tool and then incorporating the result into your review process. We started doing the work that specific way so that we could provide the lawyers a list of the links to the external references. The first attempt at this approach concluded that the entire document reference would take about two minutes to search, each one individually, and to complete all 40 references would take about 80 minutes of focused work, almost an hour and a half. Much less time than doing the search the old-fashioned way using Google. I estimated that each search would have taken about five minutes to get done using the best possible references that's doing it the old-fashioned way. So that's 200 minutes or a little over three hours of work. Using the artificial intelligence tool could maybe cut that in half.
3:00 Speaker 00 But it's still nearly an hour and a half of steady work. The benefit to saving a lawyer three hours of manual work is obvious at their well-deserved hourly rate. But then there's one more approach that's even more powerful. If you give the AI tool the entire agreement as an input and then ask the AI tool to create a document with the links to all of the external references, the AI tool can do the job in a single shot. In this instance, the entire task was reduced to four and a half minutes, in which the output was a four-page document with all of the links and all of the references to the original locations in the original document. Now, as humans, we're used to asking succinct questions. Humans can only process so much information at once. If you ask a rambling question of a human, they're likely to cut you off and ask you to get to the point of the question.
3:45 Speaker 00 I've experienced this more than once when my wife has done that very thing and asked me to get to the point. You would never imagine asking a question that involves reading 49 pages. Nobody could ever even absorb the question, let alone answer it. That's precisely what an AI tool can do and a human can't. You can ask an AI tool a question that's 49 pages long. You could ask a human to take forty-nine-page document and find every external reference in the document, perform a search to link the document to the external references. This is a fairly clear and succinct set of instructions. But the human will struggle by having to perform that set of instructions, in this case forty times. The machine is not bothered by that tedium. So the obvious question is, did the AI tool provide good references? Well, the answer is yes.
4:29 Speaker 00 In this instance, most of the references use the online resources from the law school at Cornell University. Now, I'm not a lawyer, and I certainly don't pay one on a podcast either, but frankly, I was happy with the result. It was a better result than I would have realistically performed myself. Eventually, fatigue would have set in, and some of the later external references probably would not have been as well researched or as diligently researched as the first few. The credit for all this work is not mine. It goes to my team, to Robbie Butler and our team, who's become a student of AI and how to use it most effectively, and to Autumn, our assistant, who actually did the work. By combining the thinking of the three of us, we resulted in finding a way to use the AI tool, not to get a twenty or fifty percent gain in productivity, we easily experienced
5:13 Speaker 00 fifty times improvement in productivity by changing the way that the question was asked. We effectively used the AI tool as an agent to perform the work. This is the power of AI that requires different thinking on the part of the user to truly leverage its potential. As you think about that, have an awesome rest of your day. Go make some great things happen. We'll talk to you again tomorrow.
Previous Impact of Federal Gov Shrinking Next When Things Get Stuck At The Closing Table Page updated November 15, 2026 · transcript and summary are machine-generated