Stop Using AI Like Google: Practical AI Strategies for Real Estate Investors with Michelle Hamilton

Michelle Hamilton (00:00)
one of the things that we’re doing at Answer Rocket is we are actually building out synthetic personas. So you could, if you were getting ready to, you know, pitch an investment to your bank or, you know, pitch an investment to other investors to come together with you, you can actually

actually upload custom instructions and ask your AI bot to take on those personalities and then you hit that voice mode and you tell it, I need you to interview me as this persona.

Don’t make it easy on me. We will talk until we’re done. And then you’re going to rate, how did I do? How could I have done better if this was my ultimate goal? It is amazing at taking on the persona given the chance.

Cody Crabb (02:21)
Welcome back to the Real Estate Pros podcast. I’m Cody Crabb and today I’m joined by Michelle Hamilton. Now she comes out of the commercial real estate world and now she helps firms actually make AI useful in their business. Today we’re going to be talking about how operators can use AI to multiply their output. They can improve their workflows, get some actual ROI and not waste money on the big flashy new things because I know that is a very, that is a very tempting thing nowadays. So ⁓ Michelle, thanks for joining us today.

Michelle Hamilton (02:50)
Cody, thanks. I love being here, appreciate it, and you are absolutely in a space that I am so passionate about.

Cody Crabb (02:57)
All right, so give ⁓ us just kind of a quick ⁓ overview of your background. So you mentioned ⁓ you’ve kind of pivoted now, but in the past you worked quite a while in real estate. So tell us more about that.

Michelle Hamilton (03:05)
ahead.

did.

I did. So Cody, ⁓ prior to my current role, I worked for architecture, engineering, construction, construction management teams, predominantly in the US, ⁓ really building all building types, everything from both residential to commercial. And for me, it really is all about

the built environment. How do we all live, work, eat, play, pray, whatever it might be. ⁓ For me, it is really truly understanding and appreciating how teams come together from the investment all the way through move in in real estate.

Cody Crabb (03:54)
So, ⁓ and now I’m curious, like what do you see, so you’ve got this long background in real estate and now kind of you’ve pivoted into, were chatting before the podcast and you mentioned you pivoted to AI right before chat GBT got popular, which I commented that must be the greatest timing of all of history, because that’s, you probably just got into it and then watched the whole world get wise to it, so.

Michelle Hamilton (04:14)
It was.

Cody Crabb (04:20)
Tell us about that, that must have been pretty crazy.

Michelle Hamilton (04:20)
I did.

You know what? I was lucky enough to get a very early beta version of ChatGPT about six months before the general public. And so ⁓ at the time, my youngest child who was still in college at the time, the two of us began playing with it back and forth, iterating, understanding how is it built and how can it actually impact everything from education. And at the time I was working ⁓ as the vice president of an architecture firm. And so I was applying.

some of the AI principles, practices, and strategies to business development, marketing, operations, but really thinking about how to take historic data, make sense of it, and then use AI not to replace my team, but to augment my team. so having that six-month head start meant we got to really work very closely together. And then, of course, the rest of the world began hearing about it.

Cody Crabb (05:12)
Hmm.

Michelle Hamilton (06:11)
By then I had already started consulting and sharing what I had been learning with so many of the other firms and I’m in St. Louis so with quite a few of the other firms here and I happen to sit on a women’s commercial real estate board an international board and so my ability to help other leaders in their firms think about having an AI lens to the amazing work that they were already doing.

helped me to understand that I was in the right place at the right time. And because of my business development and marketing background, I came at it really truly as a communicator. And how do you get this piece of technology and understand how to use generative AI in a very conversational manner, as opposed to thinking of it as this IT problem? I was more thinking of it as a people transformation.

problem and helping people to relax and be comfortable and then just have a conversation.

Cody Crabb (07:18)
Yeah, that’s really interesting to think about because it’s it’s it is technology but we’re almost out of the phase of even treating it like technology anymore. It’s like it’s almost like like a robot HR. Like how do you how do you how do you communicate and talk to? Yeah, so it’s kind of interesting. So, OK, I you know, when you’re going through training at a at a job and it’s like, here’s Frank, he does everything wrong. He breaks stuff and he, you know, steals paper clips and he does all that stuff. And then you have

the other guy and he really, he’s the great employee. I’d be curious to know, what are the Frank and Johnny of AI? When a company is implementing AI, let’s maybe stay within the real estate ⁓ genre here, but when someone is implementing AI, what do you see where they’re like, why would you do it that way? one thing implies, you already implied a minute ago that you don’t think it’s to replace people. I’d love for you to include that as well.

Michelle Hamilton (08:16)
Sure, so you know, I actually think of two different buckets of mistake buckets. The first one is the organizations and the leaders that say, you may not use AI at all, it is forbidden, or I’m gonna give you one tool and one tool only, I’m not gonna actually show you how to use it, and I’m not gonna show you how to use it for your job, and you’re gonna feel really threatened that I brought it in to take your job, but here you go and don’t you dare use anything else. That is one of the biggest mistakes I see.

Cody Crabb (08:46)
Just

pause, before we go into the good examples, what percentage of employers and people that own businesses do you think are like that?

Michelle Hamilton (08:53)
You know, at first, more so in the beginning, I ran into a lot of companies that said, mm-mm, no way. Now I still hear a lot of companies saying, no, there’s only one tool and one tool only, and therefore…

most people because we are an innately curious being, if the one tool that has been handed to them is insufficient or they have not been trained how to use it for their job, not just trained to use it, but how do I advance my own KPIs?

Cody Crabb (09:24)
Yeah.

Michelle Hamilton (09:28)
then they are simply going to pick up their phone and use whatever app they’re using at home to look up spaghetti recipes. And they’re gonna just start leaking your company’s data into an unapproved model because you have not invested in a people-first model in explaining to them.

Cody Crabb (09:39)
It’s a good way to look at it, yeah.

Michelle Hamilton (09:49)
and letting someone like me and like the company that I work for come in and actually listen first and understand, you know, at let’s say a construction firm, understand what are the site supers doing? What are the project managers doing? How are we thinking about their existing role?

And now let’s think about how AI can be a part of that instead of just assuming they already know how to use it. And my favorite analogy with that is you would not just send people to a job site and say, there’s a bunch of tools, build me a building. You actually have to send people with training that know how to use the tools to build a safe home, a safe multifamily, retail.

Cody Crabb (11:11)
Yeah, you might end up with a building. No guarantees

though. Like who knows what you’re gonna come out with. Yeah. Yeah.

Michelle Hamilton (11:14)
Mm-hmm.

because people were not trained to use their tools

and the trades people were not coming in with expertise about their particular roles and use cases. AI literally can be thought of in the same way. so, you know, what I’m doing now is I work for an AI company called Answer Rocket and we deliver global AI consulting and solutions really truly around the world. But I lead their division that works with

companies, company owners, and individual investors in understanding how to use AI for their job, their role, and their goals, and doing it with a human first lens, not just saying, here’s a slide deck, good luck. So that is the best analogy that I can give is ⁓ a job site.

Cody Crabb (12:02)
Yeah, yeah, yeah.

Totally, yeah, that makes a lot of sense, because it’s true with any job. you’re giving someone, like, how many times have we been through PowerPoint training or something? Like, you have to use all the features, but for some reason with AI, it’s like, you’ll figure it out. ⁓ Yeah, very much so.

Michelle Hamilton (12:13)
Yes.

again.

You’ll figure it out. And it’s dangerous to do it that way because people don’t understand

what you should and shouldn’t put in there. They are putting out garbage that does not represent your brand, your values, your particular goals because AI makes everything sound amazing. But unless you have learned how to have a conversation with it and ensure that you are requiring it to cite its sources, prove its points through hyperlink

Cody Crabb (12:51)
Yeah.

Michelle Hamilton (12:53)
you know, and using multiple bots to double check each other, then it just sounds so good. People are willing to put it out there. And I have to tell you, it’s like telling a rumor you heard at a party. It’s on you if AI lies. It is not the AI’s fault. It is on you for repeating it and not checking your work. And so that is part of the training. It is.

Cody Crabb (12:54)
you

Yeah, that’s, hallucination is a big, that’s a massive concern right now. So, okay, so

let’s kind of apply this to a real estate mindset. our audience is mostly real estate investors. You have a pretty good sense of what the day to day of a real estate investor would look like. Why don’t you give me some, a couple of good and a couple of bad examples.

Michelle Hamilton (13:21)
Yes, please.

Cody Crabb (13:35)
of a way that a real estate investor might think of AI and using AI because I guarantee everyone listening to this has either started to use it or has thought about using it. So we might as well just, you know, we might as well figure out like, how are we going to do it? How are you going to use it then?

Michelle Hamilton (13:49)
Right, I love that question. So, you know, one of the ways that I would love for your investors to sort of take a deep breath and decide, just pick one large language model and it can be chat GPT, because that’s kind of a Kleenex of AI or Gemini or Claude. To be honest with you, I am agnostic about.

Cody Crabb (14:04)
Yeah.

Michelle Hamilton (14:09)
I may have my favorites, but when it comes to people’s use cases you have to start somewhere and you need to start with a paid model. Don’t cheap out and go free. It is worth the three cups of coffee that $20 a month is going to cost in order for you to get far better results. So if we’re applying, let’s just say, ⁓ a Claude lens to that. When you are an investor and money counts and

Cody Crabb (14:23)
Yeah.

Michelle Hamilton (14:38)
counts, understanding how to ask the model to begin doing research for you is usually one of the first places that people start. But people also start with basically, like I said, using it like hyper Google and they’ll just say, ⁓ find me this type of property or help me find a family office that is going to have these dollars available.

They pretty much, I mean, if you think of AI like a smart intern, if you had a smart intern come into your investment company and you just said to she or he, here’s our 250 page investment report, give me a summary of it. They’d go, oh.

Okay, and they’d carefully read every page and they’d write you some report that had nothing to do with what you wanted. Had you said to them, I need you to read these 250 pages, here are the top five things that I need to know about. I need to know what page those happened on. I need you to predict how that is going to apply to my business goals. And then we need to write that out into both a board report, I need to write a letter to my bank advisor.

And then when you need to think about how this is going to impact my portfolio, you’ve basically told your AI bot, what I like to refer to as who, what, where, when, why, how. Give it all the information you have up front. mean, tell it your life story about why ever you’re looking for a particular, like I said, reading an investment report.

Cody Crabb (16:56)
Yeah.

Michelle Hamilton (17:02)
If you give it that much context and then you end that first prompt with, what else do you need to know from me before you get started? Your AI bot not only will understand your goal,

Cody Crabb (17:12)
that’s good too, yeah.

Michelle Hamilton (17:17)
why you’re doing this, what you’re planning on doing with it, and why it matters to driving your business forward, it is going to synthesize all of that information. And by the way, it can read 250 pages in about 10 seconds.

Cody Crabb (17:34)
Yeah, like just think of it. Imagine just being told that 10 years ago. Like what could you do with that? It’s amazing.

Michelle Hamilton (17:40)
Exactly.

But it is also then going to ask you probably three to five really great targeted questions so that it begins to understand what else do you wanna know about this? We’re just gonna use that large document analogy again. What else do you need to know about this? So if that is your beginning iteration, when you are having a conversation, just that skill set alone is going to 10x the way that you’ve been using AI up until now.

That is a really great place to start. The other place to start is understanding the tools that are available to you already pre-built into that AI model. Everyone thinks they have to invest in 10 different types of models and a specialty model for this and for that and a third party tool for this. Most of the large language models,

have probably a dozen tools that are already pre-built into them that most people don’t know about. They don’t, for the most part, come with user’s manuals. Here’s my favorite secret. Ask your AI model to tell you what are all of the tools that are pre-built into you that I don’t know about. Write me a user’s manual of how I can use you and all of your tools for real estate investing. These are my goals. These are my areas of investing interest.

Here are places where I know I’m going to need some personal assistance. Maybe here’s my website or here is my brand voice. This is how I like to represent myself in communications. Tell your AI model that bit of information about you and then have it build a user’s manual built just for you.

Cody Crabb (19:29)
Yeah, I mean, it gets to the point where you can even say stuff like, knowing what you know about me, make sure you focus on the stuff that I do and ask you about all the time. I think people do kind of think, I’m a little bit of a hobbyist myself, I’ve been big into the AI stuff. I think people think of it like Google, like you said, like a hyper-Google, it’s really targeted and things. if you really…

Michelle Hamilton (19:34)
Mm-hmm.

Cody Crabb (19:52)
If you treat it like more, it will give you way more. It’s like the better prompts you give it, it’ll multiply them by several times. Yeah.

Michelle Hamilton (19:59)
you

And always asking for more. I love that you use

the word more. You should never ever be satisfied with the first answer it gives you. You can say, that was pretty good, but here’s what I didn’t like. give me three, let’s say you’re gonna do a LinkedIn post to help highlight a new property that you’ve invested in. You’re excited about a project that’s gonna be built out or maybe you’re looking for a developer. Well, let’s say you’re building out LinkedIn posts around that. Don’t be satisfied with the first one. Say, well, that’s pretty good.

more versions of that but help me sound more like X. ⁓ wow I really like one and three. Combine those together. Give me three more. It doesn’t…

Cody Crabb (20:39)
Yeah, yeah. I really like this guy’s

posts. Can you make me sound more like that? He’s really funny. the more specific you are, the better it will be. Yeah, I totally agree with that. So.

Michelle Hamilton (20:45)
Yes.

Yes.

I think with your

investors too, Cody, and I just wanted to say one last really quick thing before I forget. Numerous AI models are actually not aware of what today’s date is. You would think that would be the most basic thing, right? I generally tell people, tell it upfront, know, today, what is, today is March 26, 2026. I only want current information because it’s built out on the, basically the entirety of the internet.

Cody Crabb (20:54)
Yeah.

Michelle Hamilton (21:19)
If you don’t tell it that you’re looking for current information as of today, it’ll bring you back great information about maybe a leader at a company or an ideal client profile, but it could be old information and you’re gonna feel frustrated by that.

Cody Crabb (21:36)
Yeah,

okay. So this has all been fantastic. ⁓ Just a couple more questions and partially just because I want to know the answers to these. What is something that someone can add to, this is a little more advanced if you’re already using AI, but what’s something that someone could add to ⁓ custom instructions for their AI that would really make a big difference? Because just kind of overall. One thing that I add, like you just said, I say look up the date every time. Google what the date is first before you do anything.

Michelle Hamilton (21:41)
You

Cody Crabb (22:05)
and then make sure it’s current. But I’m curious if you have any other little tips like that.

Michelle Hamilton (22:05)
Yes.

I absolutely do. First of all, I generally recommend to people that they name their AI bot, they actually give it a human name. We have all gotten used to talk to Siri and Alexa and everybody’s always yelling at their, Alexa, be quiet, Siri, make a call. If you name your bot and you tell it that it is its name in the custom instructions and then here is your name,

It allows your brain to relax into that conversation and mine happens to be named George. And so, and I do tell you, I recommend to people that I use AI probably 60 to 70 % of my day in voice mode or in transcription mode so that I can simply talk and mostly Cody, it’s cause my mouth moves way faster than my hands could ever take. Yeah. Yeah.

Cody Crabb (23:02)
Well then you don’t have to organize it. You just kind of, bleh, and it’ll

figure it out. Yeah. That’s, yeah, that’s an interesting thought. Yeah.

Michelle Hamilton (23:06)
Yes.

You know, and so naming it is really,

is really helpful. But that voice mode piece, think is one too that you’ll really like and your investors are going to love because, you know,

one of the things that we’re doing at Answer Rocket is we are actually building out synthetic personas. So you could, if you were getting ready to, you know, pitch an investment to your bank or, you know, pitch an investment to other investors to come together with you, you can actually

actually upload custom instructions and ask your AI bot to take on those personalities and then you hit that voice mode and you tell it, I need you to interview me as this persona.

Don’t make it easy on me. We will talk until we’re done. And then you’re going to rate, how did I do? How could I have done better if this was my ultimate goal? It is amazing at taking on the persona given the chance.

Cody Crabb (24:09)
This is all such good information. So what I would recommend to everybody, ⁓ as you’ve kind of seen here, is it really just depends on your imagination, like what you can actually use AI for. So I would recommend, like, you know, ask your AI tool of choice, like, hey, what are some creative ways I can use you for real estate investing or for whatever else? And then go, get more out of the box. What could I really use you for? And then do it again.

Now, okay, now really go out there. What could I use it for? And there’s things that I had no idea they could do that it just is like, no way. ⁓ Chat GPT, I believe, has a Zillow integration now. ⁓ So there’s things like that that it’s like, you need to explore these things. And just a little note of mine.

Michelle Hamilton (24:46)
It does.

Cody Crabb (24:56)
This is not supposed to be the Cody podcast, this is about you. ⁓ But one little note I wanted to add too was ⁓ the quote that I always hear about the jobs, know, people are worried about like, is this job or that job gonna go away? And I always say, you know, an AI is not gonna replace a lawyer, but it will replace a lawyer who doesn’t use AI. So like, so somebody that, yeah.

Michelle Hamilton (25:17)
It will. I’m the same here with your doctors, by the way. I actually ask

all my doctors now, do you use AI? And you should know I am transcribing our conversation because I now transcribe every conversation. And the doctors that say no, I will have to tell you I second guess whether they will be my doctor moving forward. Not because I think doctors are not unbelievable, brilliant human beings, but they are also human beings.

that could lean into the collective knowledge that all doctors have contributed. And I think that having an open mindset to that shows an ability to continue to grow and investors are no different.

Cody Crabb (26:02)
Yeah.

Yeah, very much so. think that’s a great way to look at it. This has been really ⁓ informative. Michelle, thanks so much for giving us all this. ⁓ If somebody is hearing this and they think they want to work with you or they want to get some tips from you, what’s the ideal person that should be reaching out to get some advice from you?

Michelle Hamilton (26:26)
I appreciate that. So the ideal person is really somebody ⁓ that is looking to truly understand how to use AI to grow their business and grow their business practices and do it in a way that they are already understanding by simply applying an AI strategy, lens, and conversation to whatever tools that they feel comfortable with. And although now I happen to work with an organization

that handles global companies, we do work with some very large real estate organizations, brokers, investors, and so for me now, that happens to be my ideal client profile, but people also can integrate with me on LinkedIn because I love.

I love chatting back and forth with people on LinkedIn, swapping stories, hey, try this, have you ever thought about that? I mean, that is just a great place to have an open dialogue and conversation and tap into someone that, you know, I love connecting the dots and sharing, so I am open to those conversations and that’s a good place to start.

Cody Crabb (27:43)
Awesome, yeah,

go connect with her on LinkedIn. I’m gonna be doing it, you might regret that, because I have so much to say. ⁓ Well, thanks for joining us today, that we’ve given us a lot of awesome stuff. And thank you listeners for hanging out with us today. If you liked what you heard, go ahead and ⁓ like and subscribe and comment and all the things. And we will see you next time, Michelle. Thanks so much one more time, and we’ll see you next time.