Latest / Net Effective / AI agents sound easy, but they aren't: inside real estate ops

Transcript

AI agents sound easy, but they aren't: inside real estate ops · Net Effective

• 0:00
  SPEAKER_01: Everybody kind of underestimates how much work it takes to use AI and then overestimates how much of AI will do for you.
• 0:07
  SPEAKER_01: Me having to pull a regrol and then go check the renovation status of every unit and manually put it in there and then figure out what we spend on it.
• 0:15
  SPEAKER_01: That process is gone.
• 0:17
  SPEAKER_01: You're putting all this information into a company that has no duty to keep it private.
• 0:23
  SPEAKER_00: That's James Goody, and I'm Jonas Bordeaux, CEO and co-founder of Dwellsey, and this is NetEffective, the Rental AI show.
• 0:30
  SPEAKER_00: I'm really pleased today to be talking to James Goody, who's a longstanding friend of mine and an incredible person in our industry.
• 0:36
  SPEAKER_00: He is the managing partner of Oak 21 Real Estate, an advisory and execution partner, working with family offices and operators across multifamily, single family, and commercial.
• 0:46
  SPEAKER_00: Before that, he was CFO of two multi-billion dollar real estate platforms, one backed by Goldman, Sachs, and Carlisle.
• 0:52
  SPEAKER_00: And he's a CPA who's helped own and operate more than 40,000 apartment units.
• 0:56
  SPEAKER_00: Welcome, James.
• 0:57
  SPEAKER_00: Really happy to have you on the show.
• 0:59
  SPEAKER_00: Thank you for having me, Jonas.
• 1:00
  SPEAKER_00: So tell me, how are you using AI today?
• 1:03
  SPEAKER_01: So I've always been super technologically inclined, shall you say?
• 1:07
  SPEAKER_01: I've always been in charge of IT for different companies, whether it's rolling out software products like Happy Companies Inspector, where you're already procure to pay or implementing doorbell management tools or camera systems and properties.
• 1:21
  SPEAKER_01: And I'm super excited about AI.
• 1:23
  SPEAKER_01: I started using it kind of slowly at first and now kind of picked up speed as the chat uh engines keep evolving.
• 1:31
  SPEAKER_01: And kind of one of the major ways that we're using AI right now is connecting all of our systems.
• 1:37
  SPEAKER_01: We're connecting QuickBooks Online, App Folio, Monday.com, and using them to pull insights on all of our properties in a way that we could never do so before.
• 1:47
  SPEAKER_00: Tell me more about those connection points, because that's been one of my biggest challenges in working with AI is trying to get everything connected.
• 1:54
  SPEAKER_00: It never seems to be as easy as I want it to be.
• 1:56
  SPEAKER_01: It's not.
• 1:57
  SPEAKER_01: And that's one of the things that's interesting about it is AI is talked about is it's a super easy to use tool.
• 2:03
  SPEAKER_01: It's very easy just to roll it out.
• 2:04
  SPEAKER_01: People are creating these six-figure businesses.
• 2:06
  SPEAKER_01: And in fact, there's an article in the Wall Street Journal about a guy in Sostalino, California that has a company that he created just with himself and AI agents making six figures.
• 2:15
  SPEAKER_01: And they talk about in the article that everybody kind of underestimates how much work it takes to use AI and then overestimates how much AI will do for you.
• 2:24
  SPEAKER_01: So getting everything to work together, it can be simple sometimes and sometimes can be a lot more complicated.
• 2:29
  SPEAKER_00: Yeah, it seems to be like the breakpoints between the systems are oftentimes where we get into trouble.
• 2:35
  SPEAKER_00: So I'm curious, across all of these different areas, you're linking all these together.
• 2:38
  SPEAKER_00: What's the goal?
• 2:39
  SPEAKER_00: What are you trying to accomplish by linking all these systems together?
• 2:42
  SPEAKER_01: I think the goal is to make life easier, frankly.
• 2:45
  SPEAKER_01: It's to get more done with what you're doing.
• 2:48
  SPEAKER_01: And I'll give you an example is in the past, people wanted to outsource.
• 2:51
  SPEAKER_01: You outsource your team, you're saving costs, you're saving time and saving staff.
• 2:55
  SPEAKER_01: And part of outsourcing, I've done this with a 30-person team of one of the companies I was with, is that you turn a lot of your people into reviewers.
• 3:03
  SPEAKER_01: So your outsource team does the work, you review, make sure it's accurate, and then you freeze yourself up to do more high-level work.
• 3:09
  SPEAKER_01: And that's exactly what I'm seeing with AI is that by using it to connect these systems together, pull data that you might have to manually pull, you've moved from a person that's actually doing the work to a reviewer that's able to take a higher level viewpoint of what's going on.
• 3:23
  SPEAKER_00: Yeah, I'm definitely seeing the same thing in my work.
• 3:25
  SPEAKER_00: Love being able to review work instead of having to do it the first time.
• 3:29
  SPEAKER_00: But I'm curious, you mentioned outsourcing there.
• 3:31
  SPEAKER_00: Do you think this is a replacement for outsourcing, offshore workers, that kind of thing?
• 3:35
  SPEAKER_00: Is that where this is going?
• 3:37
  SPEAKER_01: It depends on who you talk to.
• 3:38
  SPEAKER_01: So if you talk to all the tech CEOs as of a week ago, they all said, no, it's not replacing everybody.
• 3:44
  SPEAKER_01: But I have a 21-year-old daughter, and she went to a Giants game recently, and she was with a friend that had that was an intern in an AI company, and they stopped by 10 o'clock on a Friday night.
• 3:54
  SPEAKER_01: This company where they were teaching outsourced workers, they had them using these headsets that were recording what they were doing, and basically learning how to do their jobs.
• 4:02
  SPEAKER_01: Yeah.
• 4:03
  SPEAKER_01: And so what does that mean?
• 4:04
  SPEAKER_01: Does that mean they're replacing them or does that mean they're just helping them?
• 4:07
  SPEAKER_01: I don't know.
• 4:08
  SPEAKER_01: But I think there's a little bit of both.
• 4:09
  SPEAKER_01: You still need people to do the work and review the work.
• 4:12
  SPEAKER_01: And certainly, you know, physical jobs need people, but I think there's going to be a combination of both replacement and augmentation.
• 4:18
  SPEAKER_00: Yeah, it's such an interesting and delicate question.
• 4:20
  SPEAKER_00: I kind of went straight at it, but I know we kind of talk around these things a lot of times.
• 4:24
  SPEAKER_00: And there are certain areas where, you know, for example, with technical resources, I've never been in an organization in my entire career where there's enough technical resources.
• 4:31
  SPEAKER_00: Everybody always wants more, more IT support, more developers, all of that.
• 4:36
  SPEAKER_00: But you may not need more people at the property.
• 4:38
  SPEAKER_00: And I think that's an important thing to think about.
• 4:40
  SPEAKER_00: What can we do efficiently that provides a better experience for the resident versus where do we really need the people in order to create the best experience for folks?
• 4:48
  SPEAKER_00: So it's interesting to be able to figure out those trade-offs a little bit differently with AI in the toolkit.
• 4:53
  SPEAKER_01: Yeah, I had recent experience.
• 4:54
  SPEAKER_01: I was using one of the AI recording softwares I've used for a long time.
• 4:58
  SPEAKER_01: And I was having problems with my plan and I went to go get help, and all it was was AI agents.
• 5:03
  SPEAKER_01: You literally could not onto somebody or get a you can't even get the old, you know, you start yelling at the computer and saying, agent, agent, operator, all that took me back to was AI.
• 5:13
  SPEAKER_01: Say, you seem upset.
• 5:18
  SPEAKER_01: In property management, you know, business can break where if the clients can't actually get real life help when it comes to an issue, it doesn't matter how automated or how you know slick your systems are, they're still not going to be happy.
• 5:29
  SPEAKER_00: Yeah.
• 5:29
  SPEAKER_00: No, that's so true.
• 5:30
  SPEAKER_00: We go astray if we forget that these are people's homes.
• 5:33
  SPEAKER_00: This is a very different relationship with a product, if you will, than it is with something else.
• 5:38
  SPEAKER_00: There's a lot of things you buy a pair of shoes and they don't work out the way you hoped it, so be it.
• 5:42
  SPEAKER_00: You can kind of deal with lousy customer service and it's not the end of the world, but this is your home.
• 5:46
  SPEAKER_00: Uh lousy customer service is not acceptable in that situation.
• 5:49
  SPEAKER_00: So, how do we make sure people that we're using these tools to get better service, not necessarily cheaper service, right?
• 5:56
  SPEAKER_00: That's uh should be all of our goals in the end, is better experience, I think, for the renter and you know, ultimately a more profitable investment.
• 6:02
  SPEAKER_00: But along the way, my experience, you have a more profitable property if you have longer standing tenants.
• 6:08
  SPEAKER_00: And that usually happens.
• 6:09
  SPEAKER_00: The residents stay longer when they're happy.
• 6:12
  SPEAKER_01: And the name of the game right now is ten of retention.
• 6:14
  SPEAKER_01: When you rest flat and if somebody's least is above market, you want to keep them in place.
• 6:18
  SPEAKER_01: You don't want to give them reasons to leave.
• 6:20
  SPEAKER_01: One of the big reasons is just a lack of communication.
• 6:23
  SPEAKER_00: No, so true.
• 6:24
  SPEAKER_00: So getting specific, you mentioned connecting all these existing systems.
• 6:27
  SPEAKER_00: That makes sense to me, connecting up Folio QuickBooks, et cetera.
• 6:30
  SPEAKER_00: What are the actual tools that you're using that are AI tools in order to get these things done?
• 6:35
  SPEAKER_00: You're not a coder, you're not a software engineer.
• 6:38
  SPEAKER_00: What are you actually using?
• 6:40
  SPEAKER_01: I mean, for us, I'm using right now what's my best friend, which is I think is a lot of people's best friend, is my friend named Claude.
• 6:46
  SPEAKER_01: And that's that's often where I start with trying to figure out how to connect things and use it.
• 6:50
  SPEAKER_01: A lot of people say, don't use Claude or ChatGPT as a web search.
• 6:54
  SPEAKER_01: Like that's dumb.
• 6:55
  SPEAKER_01: You're not supposed to just use it to search for things like you would in Google.
• 6:58
  SPEAKER_01: But I disagree with that.
• 6:59
  SPEAKER_01: I I think you can use it to do kind of more of an advanced web search, meaning I might be able to search for a term and it could take me a lot of time to go through links on the internet.
• 7:08
  SPEAKER_01: But with Claude, it gives me a lot of different things I might not have thought of or dealt really deeply into all that.
• 7:13
  SPEAKER_01: You know, the hour of research I might have done can get you know knocked down to 30 seconds of Claude coming up with an answer for things for me to look at.
• 7:21
  SPEAKER_01: And so, you know, we're using some of the connectors built into the software, Claude Connectors for Excel or Cloud Connectors for Patfolio.
• 7:28
  SPEAKER_01: There's a company that um I'm the advisory board for called Ledger Systems, LDGR, and they're automating using AI, your utility payments.
• 7:38
  SPEAKER_01: So utilities and AP.
• 7:40
  SPEAKER_01: And so connecting that to your systems, having them automatically pull information, pull bills, do all the stuff that you'd have to manually do and double check.
• 7:48
  SPEAKER_01: They're doing all that in a fraction of the time and at a fraction of the cost of a lot of the traditional utility systems.
• 7:53
  SPEAKER_01: It's just one tool that we're using that's that's making life a lot easier.
• 7:56
  SPEAKER_00: That's interesting.
• 7:57
  SPEAKER_00: I was talking to a bunch of folks about utilities the other day.
• 7:59
  SPEAKER_00: It's one of the most complex areas of our business.
• 8:01
  SPEAKER_00: I think one of the most underappreciated and least understood.
• 8:04
  SPEAKER_00: Uh, the amount of complexity in that world is is staggering.
• 8:07
  SPEAKER_00: So that's really interesting about Ledger.
• 8:09
  SPEAKER_00: So, in terms of you know, actually getting net effective benefit from all of these things, where are you seeing the biggest bang for the buck in terms of time expended so far?
• 8:19
  SPEAKER_01: I think the biggest bang for the buck really is freeing yourself up to be able to do a little bit more.
• 8:24
  SPEAKER_01: Ledger Systems is a great example of that.
• 8:26
  SPEAKER_01: They're working with a property manager that we work with, and in talking with the property manager, they're saying, wow, it's amazing.
• 8:33
  SPEAKER_01: We used to have our accounting person spend all this time going through and reconciling these things and checking on it.
• 8:38
  SPEAKER_01: Now it's done.
• 8:39
  SPEAKER_01: We go in, it's done, it's run.
• 8:41
  SPEAKER_01: And so now all of a sudden we're spending time on looking at, hey, wait a minute, this will be utility trends.
• 8:46
  SPEAKER_01: Why were our utility bills higher in this month?
• 8:48
  SPEAKER_01: What's going on?
• 8:49
  SPEAKER_01: How do you compare that to occupancy?
• 8:51
  SPEAKER_01: So it's it's freeing you up from that what I was talking about earlier, this kind of higher level review work.
• 8:56
  SPEAKER_01: So you're not so much in the weeds of doing these things that actually able to review things and get insights you might on that previously.
• 9:03
  SPEAKER_00: Yeah.
• 9:03
  SPEAKER_00: So do you find all that means you can move faster, that you can process more, do more with less resources?
• 9:09
  SPEAKER_00: What's the bottom line for you on all that?
• 9:12
  SPEAKER_01: I think it makes us smarter, honestly.
• 9:14
  SPEAKER_01: I think we're doing more with less resources.
• 9:17
  SPEAKER_01: Me having to pull a rent roll and then go check the renovation status of every unit and manually put it in there and then figure out what we spend on it.
• 9:25
  SPEAKER_01: That process is gone.
• 9:26
  SPEAKER_01: That's all automated these days.
• 9:28
  SPEAKER_01: And now we can get even better insights of not only the units we've renovated or not renovated, but then get insight from the eye of which units are taking up the most time to renovate or which ones are lagging the most.
• 9:39
  SPEAKER_01: And it gives you the outliers so that you're not, you know, the 80-20 rule, you're not focused on the units that are maybe the hardest to renovate or having the biggest problems taken away from your focus for renovations.
• 9:51
  SPEAKER_01: It's giving the insight to go, hey, this is where I need to focus my efforts, and this is maybe information I didn't know.
• 9:56
  SPEAKER_00: Being able to focus is a huge thing and being able to make smarter decisions is also an enormous thing.
• 10:03
  SPEAKER_00: How do you make sure that the information's accurate?
• 10:05
  SPEAKER_00: How do you make sure that you're not dealing with hallucinations or other um, you know, kind of traditional issues?
• 10:11
  SPEAKER_00: If we can talk about traditional issues with AI.
• 10:14
  SPEAKER_00: That's been a big challenge.
• 10:15
  SPEAKER_01: I've been trying to use AI agents.
• 10:16
  SPEAKER_01: I've oneday.com has been built in.
• 10:19
  SPEAKER_01: They've been pushing them a lot, and I've been trying to use them in there.
• 10:22
  SPEAKER_01: And the problem is I'm finding that there is a lot of, it's not even garbage in, garbage out, which is a kind of a data issue, data, you know, a data issue, but it's more of an issue of you're having to be so specific as to what you're trying to get the AI to pull.
• 10:35
  SPEAKER_01: So I want you to pull these columns from this spreadsheet and from this other spreadsheet or this, you know, this other board in Monday terms.
• 10:42
  SPEAKER_01: You're trying to pull from all these different information and summarize it into a new board or a new report for you so you can get all the insights if you want.
• 10:49
  SPEAKER_01: But I am finding a lot of time it's pulling the wrong column or didn't quite understand what you're looking at, or you know, the hallucination issue with sometimes just what it came up with just doesn't make sense.
• 10:59
  SPEAKER_00: No, I hear you.
• 11:00
  SPEAKER_00: It can be really hard to judge that, both the hallucination as well as the clod slop or, you know, whatever you want to call it, where you know the output is just kind of crap.
• 11:08
  SPEAKER_00: That has happened to me a lot.
• 11:10
  SPEAKER_00: It's something I've been wrestling with because it really does force a higher level of review than you might have to do had a human put that work together.
• 11:17
  SPEAKER_00: Now, it might have taken a human two or three days to do a report that now takes two or three minutes, but you have to review it in a very different way.
• 11:23
  SPEAKER_00: Uh, I feel like there's not the same guarantee of quality.
• 11:26
  SPEAKER_00: And the types of mistakes that the AI make are very different and sometimes much more concerning than the types of mistakes humans might make.
• 11:34
  SPEAKER_01: Well, you know, the other big issue right now, too, is just LinkedIn just came out with their AI slot button, basically, which you can tag any posts as being AI slot.
• 11:42
  SPEAKER_01: But it's it's super real.
• 11:44
  SPEAKER_01: You know, I've been posting on LinkedIn lately and I enjoy it.
• 11:47
  SPEAKER_01: And I was using the AI to help me write posts.
• 11:50
  SPEAKER_01: It helps you flush out ideas, helps you refine ideas, but it's really readily apparent, not only with the dash ends and then you know, not this, but that, or but here's the truth kind of thing.
• 11:59
  SPEAKER_01: But just a lot of writing looks very AI, and I think it takes away from the almost validity of the information.
• 12:05
  SPEAKER_01: You can have a great idea and spend a ton of time developing an idea.
• 12:09
  SPEAKER_01: And then when you put it out there and reads too much like AI, I think it just sounds like something that you just came up with for the moment.
• 12:15
  SPEAKER_00: There is a real AI slop problem, and I don't mean to downplay that.
• 12:18
  SPEAKER_00: I think we all need to be uh cognizant of that, and I'm I'm happy LinkedIn introduced that button.
• 12:23
  SPEAKER_00: But I also feel like it's kind of the new troll move, if you will, is just to label everything AI slop that you don't like.
• 12:29
  SPEAKER_00: So I'll be interested to see how that works on LinkedIn.
• 12:31
  SPEAKER_00: I do some uh playing around with Reddit from time to time.
• 12:34
  SPEAKER_00: And basically anytime anybody doesn't like something, they just call it AI slop and make fun of it and feels a little toxic to me, but so be it.
• 12:41
  SPEAKER_00: We'll see how that goes.
• 12:42
  SPEAKER_00: I my own pet belief is that when an argument is relatively well developed and it's relatively thoroughly described and thought through, it can be really hard to tell it apart from something produced by an AI and something produced by a human.
• 12:56
  SPEAKER_00: I think, you know, something quick and sloppy is obviously not AI, and something that is true AI slop is obviously AI, but a lot of things are in the middle.
• 13:04
  SPEAKER_00: And I don't know that I can tell the difference between them and know which one's which.
• 13:08
  SPEAKER_00: So it'll be interesting to see how that develops in time.
• 13:12
  SPEAKER_00: So I'm curious, you you talked about using Claude a lot.
• 13:14
  SPEAKER_00: Do you use ChatGPT's products or any other AI models?
• 13:17
  SPEAKER_01: I started off with ChatGPT just because that was the one that everybody was using.
• 13:22
  SPEAKER_01: Used it a bunch, and then I tried out Claude as I was reading, people liked it better.
• 13:26
  SPEAKER_01: So I figured I'd try myself.
• 13:28
  SPEAKER_01: And I think there's a you know, there's a gap in switching, it's a whole problem.
• 13:31
  SPEAKER_01: And I was actually sharing uh Chat GPT account with somebody, and they decided to delete a bunch of the chats.
• 13:37
  SPEAKER_01: I don't know what I friend of mine.
• 13:39
  SPEAKER_01: We're both in so I was like, okay, I'm getting my own account now.
• 13:42
  SPEAKER_00: Yeah.
• 13:42
  SPEAKER_01: So I tried Claude out, and I've been using it for several months now.
• 13:45
  SPEAKER_01: And it's almost difficult because you don't want to go back to using, you get used to it and get all your information now.
• 13:50
  SPEAKER_01: You don't want to try out the other one.
• 13:51
  SPEAKER_01: And maybe hey, maybe ChatGPT is better now.
• 13:53
  SPEAKER_01: And I know there's a bunch of other models out there too for people that are a little more of a you know AI snob, but you know, that's why it's been working.
• 14:00
  SPEAKER_00: No, it's it's such a funny thing.
• 14:01
  SPEAKER_00: I I found the same thing.
• 14:02
  SPEAKER_00: It's so hard to go back and forth because you get used to the form factor of one and you build up kind of a knowledge base within one of them.
• 14:09
  SPEAKER_00: But at the same time, I I can never tell who's got the better system.
• 14:12
  SPEAKER_00: And I don't know who's gonna win in the end.
• 14:14
  SPEAKER_00: So I don't want to find myself in a situation where I end up committed to a system that isn't the right answer in the long run.
• 14:21
  SPEAKER_00: But it's really hard to do that.
• 14:22
  SPEAKER_00: I keep trying to switch back and forth, but I end up using Claude most of the time anyway, because that's where I'm most comfortable on.
• 14:27
  SPEAKER_00: Uh even though it seems like maybe OpenAI is winning right now.
• 14:31
  SPEAKER_00: I don't know.
• 14:31
  SPEAKER_00: We'll see.
• 14:32
  SPEAKER_00: You you have an opinion on who's winning in that battle?
• 14:35
  SPEAKER_01: The problem is it's you know, it's like, do you love your Tesla and do you hate to be on risk?
• 14:39
  SPEAKER_01: I mean, that's common you know, we're here in the Bay Area.
• 14:41
  SPEAKER_01: That's a very common thing where you have the you know pre-Elon going for the sticker on Tesla.
• 14:46
  SPEAKER_01: I think it's the same kind of thing where you have, you know, Sam Altman is very much the evil empire right now, and you know, it's portrayed as a bad guy in the media.
• 14:53
  SPEAKER_01: I don't, you know, I don't know the truth, but you know, Anthroopic is very much portrayed as the altruistic company right now.
• 14:59
  SPEAKER_01: So just from a you know, what feels good standpoint feels like Anthropic is, you know, at least looking out for us a little bit, not trying to sell you know all of our information.
• 15:08
  SPEAKER_01: And maybe OpenI is doing that, maybe they're not, but you know, they both probably are.
• 15:12
  SPEAKER_00: No, that's true.
• 15:13
  SPEAKER_00: You know, it's funny.
• 15:14
  SPEAKER_00: I I was listening to another podcast, How I AI, and in a review of one of the latest models, she was talking about the fact that Anthropic has internalized that so much in its model that it's now like apologizing for everything and trying to be like over the top, almost making you uncomfortable with using AI.
• 15:31
  SPEAKER_00: Whereas, you know, the the Chat GPT model, it's the equivalent sort of frontier model, is just very direct.
• 15:36
  SPEAKER_00: You know, this is just the way it is.
• 15:38
  SPEAKER_00: And, you know, if you don't like it, move on, kind of thing.
• 15:40
  SPEAKER_00: So it's funny to see the personalities manifesting in the product today.
• 15:43
  SPEAKER_01: Well, and that's even when you're doing your chats or your or your projects or anything else in AI, are you still using please?
• 15:48
  SPEAKER_01: And thank you.
• 15:49
  SPEAKER_01: When you're talking to your AI, I still do.
• 15:51
  SPEAKER_01: I just it's habit.
• 15:53
  SPEAKER_01: Thank you.
• 15:53
  SPEAKER_01: That was great information.
• 15:54
  SPEAKER_00: So I'm talking to a computer, but you know, it's no, it's uh I I I've kind of decided that AI is um is a team member for me, and I treat it like I treat all my other team members.
• 16:04
  SPEAKER_00: And yeah, that means please and thank you.
• 16:06
  SPEAKER_00: And the one that I I keep catching myself on is asking it to do things, which is a little ridiculous.
• 16:10
  SPEAKER_00: But I've always found I get people like to be asked, they don't like to be told.
• 16:13
  SPEAKER_00: But you know, same doesn't really apply for AI.
• 16:15
  SPEAKER_00: So I'm I'm probably wasting a lot of keystrokes with that habit.
• 16:18
  SPEAKER_00: But that's funny.
• 16:19
  SPEAKER_00: So talk about the future and where this goes.
• 16:21
  SPEAKER_00: You know, what are you planning to build?
• 16:23
  SPEAKER_00: What are you excited about building going forward?
• 16:25
  SPEAKER_00: Places where you're gonna find more advantage for your business.
• 16:29
  SPEAKER_01: I mean, for me right now, I think the thing that I really want to learn more about is agents.
• 16:33
  SPEAKER_01: Yeah.
• 16:33
  SPEAKER_01: Everybody talks about them.
• 16:34
  SPEAKER_01: They talk about how easy it is to build, it's right there in its agents or in in everything.
• 16:39
  SPEAKER_01: It's in COD or in Monday or anything else.
• 16:42
  SPEAKER_01: And I feel like certainly for me personally, I'm not utilizing them well enough.
• 16:45
  SPEAKER_01: It almost seems like the analogy of the people that don't use AI at all or that are technophobic where they don't really know where to start.
• 16:51
  SPEAKER_01: You know, it's super intimidating.
• 16:53
  SPEAKER_01: I have, you know, family members and senior executives that were on the same team that I've worked with in the past that were scared to download an app and put a login and do it just because it seems so challenging.
• 17:05
  SPEAKER_01: And I feel like that's a little bit that way with agents because it's easy, right?
• 17:09
  SPEAKER_01: You can start one right now, it takes two seconds, you can get your own agent, but then you know how much of the information is correct.
• 17:14
  SPEAKER_01: And then the other thing that I worry worry about as a you know, CFO, COO, one of the product companies I was with when we had different phishing scams.
• 17:22
  SPEAKER_01: And I was a bank officer, I was a national director of real estate for First Republic Bank, and we have a lot of training on that.
• 17:28
  SPEAKER_01: And you're putting all this information into a company that has no duty to keep it private.
• 17:34
  SPEAKER_01: In fact, they're using your information to train its models and create its own database that they're going to use for whatever they're using it for.
• 17:42
  SPEAKER_01: And there was a lot of talk about this, probably maybe a year ago, about hey, you're not supposed to upload your company financials at a private company into AI because then all that information is now in the public realm, basically.
• 17:54
  SPEAKER_01: So if somebody else is searching for your financials, they can find it now, whereas they never could before.
• 17:59
  SPEAKER_01: So there's a lot of questions and concerns I have about that.
• 18:02
  SPEAKER_01: I'm not gonna connect my personal Gmail account to you know any of these systems, no matter what, because it has anywhere from income tax information to personal medical data, things like that.
• 18:11
  SPEAKER_01: And the other issue I see as it relates directly to real estate is this whole back and forth right now between real page and YARDI.
• 18:17
  SPEAKER_01: And YARDI's decided the open data system is the way they're gonna go.
• 18:20
  SPEAKER_01: And real page says no, we want to control our data.
• 18:23
  SPEAKER_01: And what does that mean for the customer?
• 18:24
  SPEAKER_01: And what does that mean long term?
• 18:26
  SPEAKER_01: And I'd use an analogy of beta Betamax versus VHS.
• 18:29
  SPEAKER_01: If anybody remembers that, it's uh two different formats in the 70s.
• 18:33
  SPEAKER_01: One was gonna win out, right?
• 18:34
  SPEAKER_01: One was gonna be the dominant one and ended up being VHS, got more widely adopted.
• 18:39
  SPEAKER_01: And so it's this kind of same question here.
• 18:41
  SPEAKER_01: Is it this model of all your data is going to be public anyway?
• 18:45
  SPEAKER_01: Which I I hear from you know from people and who cares?
• 18:47
  SPEAKER_01: And that's why I still say to family members, don't post your birthday on Facebook.
• 18:51
  SPEAKER_01: Why you're just giving you just making it that much easier for hacker?
• 18:55
  SPEAKER_01: And then when your mother-in-law comments on it, then they have a mother's maiden name, that type of thing.
• 18:59
  SPEAKER_01: Are we just too enthusiastic about AI right now?
• 19:02
  SPEAKER_01: And then a lot of people are willing to just put all this data out there, or is it just inevitable, right?
• 19:06
  SPEAKER_01: Is it that it's all gonna be there anyway, so what does it really matter?
• 19:10
  SPEAKER_00: That's such an interesting thing, the information getting public.
• 19:13
  SPEAKER_00: To what degree can we trust Anthropic or OpenAI when they say that that they're not gonna use it to train their model, that they're not gonna make it public?
• 19:22
  SPEAKER_00: Um, that's something that I worry about.
• 19:23
  SPEAKER_00: You know, I have all the correct settings set, I think, so that they're not making anything that I'm doing public.
• 19:28
  SPEAKER_00: But I worry that that may not turn out to be the case, or there's something that they might be doing that I wouldn't like.
• 19:34
  SPEAKER_00: And then, you know, what if those things get out?
• 19:36
  SPEAKER_00: Uh, what if people do find out?
• 19:37
  SPEAKER_00: And I often think we keep things secret that we might not need to keep secret just out of privacy preferences.
• 19:43
  SPEAKER_00: So does it really matter?
• 19:44
  SPEAKER_00: Is anyone actually looking for the performance of this real estate property or that technology company when they don't have a vested interest in it?
• 19:51
  SPEAKER_00: Maybe, maybe not.
• 19:52
  SPEAKER_00: I don't know.
• 19:52
  SPEAKER_01: Yeah, that's a key issue, but who cares, right?
• 19:55
  SPEAKER_01: Does somebody really to really care about this single family rental that we're managing in you know in Cleveland?
• 20:01
  SPEAKER_01: Probably not.
• 20:02
  SPEAKER_01: But by the same token, the last four days in a row, I've been getting pinged on my phone all the time with this is your code to resend your password for Venmo, Cash App, and uh you know what other service I'm using, and I know somebody's out there, and it has been happening every day.
• 20:16
  SPEAKER_01: I deleted Cash App for my phone because I don't use it very often.
• 20:19
  SPEAKER_01: I'm still using Venmo a lot.
• 20:21
  SPEAKER_01: I changed all my passwords for all the programs I use, but I'm still getting the messages and it plays into that.
• 20:26
  SPEAKER_01: How much are hackers using AI now too to pull information on you?
• 20:30
  SPEAKER_01: And if you have stuff that's in there that's personally identifiable information or information on your company or family members, are they getting that?
• 20:36
  SPEAKER_01: I don't know.
• 20:37
  SPEAKER_00: Yeah, we could do an entire episode or many, many episodes on just the the privacy risks around you know theft of information.
• 20:44
  SPEAKER_00: I've uh I've always worried about identity theft, and I know people have tried to steal my identity as well.
• 20:49
  SPEAKER_00: Such a frustrating thing to have to deal with.
• 20:51
  SPEAKER_00: It's scary, what's out there in the world and what people are trying to do.
• 20:54
  SPEAKER_00: So uh I hope nothing goes ari with uh Venmo and everything else for you.
• 20:58
  SPEAKER_00: Hopefully you've got that figured out.
• 21:00
  SPEAKER_00: So, what are you doing in your personal life with AI, James?
• 21:03
  SPEAKER_01: I probably use it personally as much as I do for business.
• 21:06
  SPEAKER_01: You know, what I was talking about earlier that there's this whole notion that you're not supposed to use it as a search engine.
• 21:12
  SPEAKER_01: But I mean, I that's exactly what I do use it for a lot because it used to be when you search for something, I'll give you an example, we're going to a wedding up in Healdberg in two months.
• 21:21
  SPEAKER_00: Yeah.
• 21:22
  SPEAKER_01: And so it's easy for me to go look up some hotels, you know, that type of thing.
• 21:27
  SPEAKER_01: But and it might take me half an hour to do that and then go look at availability or rates.
• 21:31
  SPEAKER_01: It's a whole lot easier for me to go to Claude and say, here's the dates we're going to be here, here's the location, here's some hotels that I know about.
• 21:38
  SPEAKER_01: Go ahead and give me pros and cons of each and give me a list.
• 21:42
  SPEAKER_01: And then that comes back to me in a minute or two minutes versus half an hour of searching the web and going on the different websites and looking at it.
• 21:48
  SPEAKER_01: But what I do then do is take that information to the accuracy of the information, is I'll take that for my recommendations and then go myself to the website for the hotel and book it myself.
• 21:58
  SPEAKER_01: I'm still not not there yet with having it book everything for me.
• 22:02
  SPEAKER_00: That's actually one of my favorite applications, too, is travel planning.
• 22:05
  SPEAKER_00: I did a trip to LA with my son, and he and I sat down with ChatGPT and basically had it plan the whole thing for us.
• 22:11
  SPEAKER_00: And, you know, at the end, we got a whole itinerary and everything organized by traffic distance, you know, with real-time traffic and everything.
• 22:17
  SPEAKER_00: It was amazing.
• 22:18
  SPEAKER_00: But right there with you.
• 22:19
  SPEAKER_00: I didn't have it actually book anything uh for me.
• 22:21
  SPEAKER_00: I'm still afraid to give it my credit card or anything like that.
• 22:24
  SPEAKER_00: That feels like a bridge too far somehow, but great use case.
• 22:28
  SPEAKER_00: One of the things I always want to do is recognize great operators out there.
• 22:31
  SPEAKER_00: Uh, you mentioned you've been working with some folks.
• 22:33
  SPEAKER_00: Who would you like to call out?
• 22:34
  SPEAKER_01: You know, there's a guy I've been working with for the past probably year and a half uh named Nick Virovek.
• 22:40
  SPEAKER_01: He's the president of Fisher Asset Management in Cleveland.
• 22:43
  SPEAKER_01: And when I talk to Nick, he's very experienced with property management.
• 22:47
  SPEAKER_01: You know, he's running, he's the president of a property management company, but he's also super progressive and tech forward when it comes to property management.
• 22:54
  SPEAKER_01: So he's always willing to try different things.
• 22:57
  SPEAKER_01: And it's great when you know when I talk to him, I'm saying I'm the fellow prop tech nerd because I'm learning about guitars and property technology, which I'm super passionate about.
• 23:06
  SPEAKER_01: But doing things like when we were even having problems with financial statements, we couldn't figure out there's a somebody had posted something in a prior period common accounting thing in in property management.
• 23:16
  SPEAKER_01: You're supposed to close a month and then nothing's supposed to change.
• 23:19
  SPEAKER_01: So the numbers don't change.
• 23:20
  SPEAKER_01: Well, they'd sent us some financials, the numbers change from the prior months.
• 23:23
  SPEAKER_01: So our accounting team was having trouble reconciling it.
• 23:26
  SPEAKER_01: And in the past, that used to be an exercise in Excel formulas and you know, hunting through GLs and all that to try to find the difference.
• 23:34
  SPEAKER_01: But for him, he he took a GL, dropped it, you know, whichever program he's using, had analyzed the difference in account balance ledgers versus the original ones that were sent, and boom, came out with numbers.
• 23:44
  SPEAKER_01: And so there's also somebody that piloted ledger systems that's been using them for AP analysis.
• 23:50
  SPEAKER_01: And then I've talked to him about a couple of other products.
• 23:52
  SPEAKER_01: And he's very into like how can this help my company and help us get more capacity so that we don't have to hire a bunch of other people and we actually get the ability to analyze and review things instead of being stuck in busy work.
• 24:05
  SPEAKER_00: Yeah, that's fantastic.
• 24:06
  SPEAKER_00: Good for Nick.
• 24:07
  SPEAKER_00: Thanks for mentioning him.
• 24:08
  SPEAKER_00: I love, love, love, love using AI with financials.
• 24:11
  SPEAKER_00: You know, just having it be able to pour through things incredibly quickly is just amazing.
• 24:15
  SPEAKER_00: So good for him for using it for that.
• 24:17
  SPEAKER_00: Well, James, it's been such a pleasure to have you on.
• 24:19
  SPEAKER_00: Thank you so much.
• 24:20
  SPEAKER_00: Appreciate you taking the time for this and appreciate all your thoughts on everything uh AI and property management related.
• 24:26
  SPEAKER_00: For everybody out there, this has been NetEffective.
• 24:28
  SPEAKER_00: We have new episodes weekly.
• 24:30
  SPEAKER_00: If you're running rentals and figuring out AI, subscribe wherever you listen.
• 24:33
  SPEAKER_00: And if you'd like to come on the show, drop me a note and let me know.
• 24:36
  SPEAKER_00: Looking forward to it.
• 24:37
  SPEAKER_00: Thanks for listening.