The AI Paradigm Shift | S4E3

In this episode of the A.CRE Audio Series, Spencer, Sam, and Michael discuss the evolving landscape of commercial real estate in a world where AI is becoming extremely prominent. They explore the real impact of AI on business and whether all the publicity surrounding it is justified. From AI-powered Excel add-ins to interview preparation coaching, the team at A.CRE shares their experiences in harnessing the power of AI to make significant strides in the field.

Join the conversation to learn about the AI paradigm shift and how it’s changing the way the industry operates.

The AI Paradigm Shift

Or Listen to this Episode

Resources from this Episode

Episode Transcript

Sam Carlson (00:00):

Welcome back. We’re in the Audio Series, season four. Here’s what I love about this season. We’re having conversations here, and then they’re just going to spawn into really cool ideas and topics. So, let’s get back with Michael and Spencer. You guys, what were we talking about? We’re talking about what’s new in CRE, but it’s more than that. It’s more specific than that.

Spencer Burton (00:20):

Yeah, a lot of hype. Well, I say hype, lots going on with generative AI, right? So these natural language models, ChatGPT, GPT-4, GPT-3.5, OpenAI‘s natural language models, but also others, the image generation models, Midjourney, DALL-E, the things that Adobe’s doing with image completion, image expansion, but also sound. So all these things are happening and the question that I think is on a lot of people’s minds, a couple of questions. The first is, is it hype or is it real? I had someone email me the other day, it was interesting. He started by saying, “Hey, Spencer, great to hear from you…” or “It’s been a little while. I’d noticed you’ve gotten in on the hype around AI,” and I thought, “That’s interesting that he said that because he’s not necessarily wrong.” Hype is … A lot of people talking about it, right? The question is, is the hype real? Is it like autonomous cars where everyone is talking about it for a while and it actually didn’t turn out to be much of anything? Or is this hype that it will have an impact immediately?

Sam Carlson (01:35):

So does it feel like … Sorry.

Spencer Burton (01:37):

Yeah, go, go.

Sam Carlson (01:40):

When people say hype, I think of all of the craziness that came with crypto, and everybody was getting in and crypto was going to change the world and this, that and the other. And it’s not that the technology there that they’re talking about, there isn’t really good stuff there, but it came in with such a force and there’s these people coming in that said, “Oh, it’s changing the world.” And it was half-baked. In many cases, half-baked. I mean, I don’t have an answer … Hold on. I don’t have an answer other than when people start talking about AI, I think they’re jumping ahead of what AI is today and what it has the potential to be, and the people who are learning or adapting to it and being able to use that as a tool instead of a replacement for something, I think that’s what …


So I don’t know that it’s hype necessarily, I think the frame of mind that you’re … Okay when you come and you say, “Is this going to replace everything or is it going to complement everything?” To me, that’s the misnomer here. I don’t think it’s necessarily hype, I just think it’s misunderstood.

Michael Belasco (02:56):

Yeah, I think the difference, when you talk about crypto and you talk about autonomous cars, these were things that were both from the technology and opening the hood and seeing what’s inside. It was out of people’s hands. When you talk about crypto, you had to sit and listen to presentations. When you talk about autonomous cars, none of us know how that technology works. And with AI, what’s happening here, and not to say that we know how the technology works, but it’s being put directly in the lay person’s hands, right? This is a technology where people are using this in use cases today. It’s not something that’s theoretical and you don’t understand it and somebody has to explain it to you. You show up to work or you show up to try and do something. Even for entrepreneurs, especially for entrepreneurs, you can take these tools today and you essentially have a company working at your fingertips, which is something you didn’t have before.


And as I was telling you guys before, and I’ll say this now, I’m exploring developing this hobby side website that I’ve just been poking around with and I have a company of probably five people that I couldn’t have had before. I have a graphics designer, somebody writing content for me, and yet I’m curating all of it and you have to learn how to work with it, but it’s amazing. I wouldn’t have hired these people in the first place, but now all of a sudden you have the capability to utilize these tools and that permeates into every industry. It gives access to everyone. Now, I want to add one piece to that is that there is a positioning that people are taking around that in this, which is some people have fear about it and then others see an opportunity. And I think it’s about the more you know and what you can see, and there is legitimate fear.


I remember Adventures in CRE before ChatGPT came out probably four or five times a day. Tell me if you got the same thing. I had somebody emailing us asking to do copy. Can we do copy for you? Can I write a piece of content for you? That’s almost disappeared overnight.

Spencer Burton (05:06):

Well, let’s come back to this idea of hype. You know what, again, when I think of hype, it’s an abundance of conversation around a topic. And hype may be real, founded in something real, and other times hype, less so. And I think really what you’re saying, Michael, is this is hype that actually is paired with a tool that people are actually using versus autonomous vehicles, there was this hypothetical opportunity in the future and all of us were talking about how are autonomous vehicles going to change the world, and in particular real estate, how is it going to change real estate use? And since it hasn’t materialized yet, and I imagine it still will be, I mean, billions of dollars have been invested in it. But this is a tool that overnight, seemingly overnight, except for those that were in the beta of GPT-2 and so forth, but that seemingly overnight all of a sudden all of us have access to a tool, an increasing number of tools.


I mean, we always think of ChatGPT, but that’s only one of thousands of tools that have been produced and created over the last 90 days related to generative AI that is changing, is changing in real time is changing how we do things in commercial real estate. And that’s hype that is real.

Michael Belasco (06:32):


Sam Carlson (06:32):

All right, let me jump in here. So we’ve been talking on this trip, driving in the car, about different ways that we want to be using this at Adventures in CRE throughout the site, throughout our accelerator program, different things we’re working on. Let’s shift the conversation that way because I think … So my take on AI today is that it is not a replacement, it is a co-pilot. And in the hands of the right person, the right prompting, the right understanding, and a person who really understands subject matter of whatever it is, commercial real estate in this case, it can accelerate so many things. So we’ve been talking about a lot of ideas, but how is this going to change real estate? But let’s talk about how… I mean, how are we implementing it in the site first and then that can evolve to how that might adapt to real estate companies, commercial real estate companies?

Spencer Burton (07:28):

Well, I think the most tangible example is our Excel for CRE Excel Add-in. So for those who are unaware, we released, I don’t know, a couple months ago, a free add-in, completely free. In fact, we don’t even accept pay what you’re able on our ad, and it’s completely free. The code is open source. But what’s cool is it’s an add-in that is almost entirely created thanks to ChatGPT. Michael and I have a cursory understanding of VBA enough to give back and create modules and know what to search for on the web. But we are not VBA developers, and we saw a need to develop a custom Excel add-in for commercial real estate professionals. And pre-February 2023, if we wanted to do that, we’d have to find a developer, VBA developer, hire that person, pay them some number that it’s less about the price, the cost, and more about the time.


And for those who’ve seen some of our videos, this isn’t new to you, but that iterative process of communicating to a developer what we want, having the developer go develop an MVP, a minimum viable product from what we want, come back to us, us say we want these changes, they go back, and you go back and forth, back and forth to ultimately have a product that, and I say product loosely, because it’s a free tool that we just give away, but a product that is actually useful and that could take weeks or months. With ChatGPT, the first version we literally released after two hours of time.

Michael Belasco (09:17):

And the truth of the matter is it never would’ve gotten done. It’s not something we would’ve put effort or resources into.

Spencer Burton (09:23):

That’s right. No, that’s exactly right. And let alone give it away for free, right?

Michael Belasco (09:27):


Sam Carlson (09:27):

It’s interesting. What about companies? So we have these different positions within a real estate company, asset manager, analyst, whatever it is. I’m sorry, analysts, associates, so on and so forth. How will this complement, replace, use, whatever adjective you feel is most appropriate, how’s this going to take effect in companies now, like this calendar year, 2023?

Spencer Burton (09:59):

Well, I wish we hadn’t have gotten away from A.CRE yet. There’s a lot of things where we are … Sorry. Yeah, I mean, you’re getting a bit ahead of yourself.

Sam Carlson (10:10):

Listen, you guys asked me to kind of guide the conversation and I started thinking, I’m like, “How’s this going to…?” But you’re right. Let’s get…

Michael Belasco (10:16):

There was a natural pause. How is Adventures in CRE going to use this? Okay, Spencer.

Spencer Burton (10:20):

So that is one use case. Michael, what’s another use case at A.CRE?

Michael Belasco (10:25):

Yeah, I mean, a critical thing we do at A.CRE is, and one of the ways we add value is teaching people how to prepare for job preparation. So we recently released a way that you can get your own personalized interview prep coach through prompting ChatGPT. So we put out this post and you basically go through it, it tells you how to prompt ChatGPT, you tell what position, you give it your experience, and it follows the prompts. And then it interviews you, and not only does it interview you, it will give you feedback about your answers, how to polish up until you get as sharp as you need to be. So you have this free resource now where you can train and prepare for interviews. So that’s another thing that we’re doing at A.CRE that would’ve never … We couldn’t have imagined doing that prior to having AI. So Spencer, okay, get back to you.

Spencer Burton (11:19):

Well, some context to that. So we do these interview bootcamps with schools, and those are still important. I mean, you can’t get everything from that, but this is a new way in which to prep for interviews. So prior to that, you’d have to find a human, you’d go back and forth, and there’s the nonverbal cues that you can’t get from ChatGPT. But what you can get is the back and forth, the feedback you can practice at least in a text form, and eventually I imagine, in a video, in a verbal form. In fact, there are tools now, if you’re unaware, I think they’re called voice coaches or something where you speak to the voice coach, you go back and forth, and then it will help you clean up the ums and the ahs, will help you be more concise with your words and will coach you on those things.


And so to Michael’s point, we’d be doing these interview bootcamps, well, you can do a similar version of that, but with ChatGPT. Or Bard. Bard is Google’s version of ChatGPT. And there’s a million of them now, not a million, but there’s a bunch of…

Michael Belasco (12:31):

Have you worked with Bard at all or really dove into that? I haven’t gotten into it that much. Have you any experience?

Spencer Burton (12:40):

Yeah. First in mind, first in right. And so ChatGPT, GPT-4, the latest OpenAI model that ChatGPT uses, I think, is still the superior option, but Bard has made some good steps and it’s free. So to get access to GPT-4, you have to be a ChatGPT plus member, 20 bucks a month or whatever it is. Bard is entirely free. I think it’s bard.google.com if you’re curious, but…

Michael Belasco (13:08):

Yeah, on the Adventures in CRE side, I’m going to kick this to you, Sam. And this relates loosely to, and this is marketing and it’s relevant to commercial real estate as well, particularly in the brokerage space. There’s a lot of video and imagery, but particularly in video, there’s AI that’s come out recently that’s revolutionized video editing. And you’re talking about in A.CRE, you do building tours with video, you do flyovers with drones and things like that. So maybe speak to some of the stuff we’re doing at A.CRE with video editing.

Sam Carlson (13:53):

There’s a couple tools. So, Opus. Opus One, or Opus something, they have … Well, first of all, so I use Descript in my video editing and Descript is cool because you’re talking about the ums and the ahs. You can click one button, it’ll identify all the ums and the ahs, and you can just go bloop and they’re gone. And when it makes the cuts, so when you make a cut in a video, instead of it being an abrupt cut, they transition and it’s smooth. And it’s almost like you can’t detect the ums and the ahs.

Spencer Burton (14:24):

Yeah. In fact, if you’re listening to this right now, there’s a whole bunch of ums and ahs, and you didn’t even get any thanks to Sam and his…

Michael Belasco (14:32):

Does it smooth out the video also, like the video transition when you’re watching it, when you edit out the ums or does it actually do a…?

Sam Carlson (14:39):

Yeah, so it does all of that. So it transitions video, it transitions the audio, which is a big one. You can train it on your voice. You can upload, I think they request 15 to 20 minutes of audio. You upload that and then, for example, you say one word when you meant to say a different word or you miss, whatever, you can go in and you can type. So instead of going into the video file and splitting the video clip, you don’t do that. You go into a document and you identify the text and you highlight it and you delete it, or you highlight it and you write in what you want and it will put that in there.

Michael Belasco (15:25):

That’s insane.

Sam Carlson (15:28):

And they’re doing so many things. I mean, there’s somewhere if we wanted to read a script, and if you’re looking down, your eyes are looking down, your eyes can look down and you can read it, but it will make your eyes look up. And so there’s so many cool things. I mean, video editing, the time it takes to identify … That one software I was telling you about, Opus, it can identify how effective it is. We’ll have to test and see, but which clips have the potential at percentage-wise to go viral. And you can upload 20 minutes of YouTube footage and it will identify 10 shorts, and you can then use that for content creation. And content’s a big piece of what’s happening now.

Michael Belasco (16:13):

So I’m thinking about the … Let me get my microphone close here. The entrepreneur starting out, let’s say you bought your first six unit multifamily building in wherever you are, and you need to market this thing now. You got your first building, you don’t have a team, maybe you don’t have the money yet to be outsourcing that, you can essentially create a marketing video. You can go and film it, maybe you still need some assistance, but effectively you can roll this up, put it into this platform, and you don’t need to be a video editor. You could essentially, and maybe it’s not there yet, but hypothetically you could edit a video by marking up and editing a Word Document. Is that essentially how this goes?

Spencer Burton (17:03):

So you’re describing tools that make a solopreneur-

Michael Belasco (17:08):


Spencer Burton (17:13):

… much more efficient. So rather than having to build a team of people around them. But this is equally being used by large organizations to be better, to do things faster. My sister’s a great example. So she is just a master copywriter, and right now she leads a team of content … She runs a content marketing division for a big company. And her and I were having a conversation the other day about how she’s using these tools and she says, first off, she doesn’t really need content writers anymore, or at least not to the extent that she did before. She will write with AI and she’ll prompt it what she needs and it will get 75% of the way there.


And her and a much smaller team can do that. But on top of that, they can then produce micro pieces of content. And again, how does this relate to real estate? I mean, if you’re any sort of a service provider, if you’re not the tenant and you’re not the investor, in between, you’re a service provider. Every one of us are a service provider, unless you’re the tenant or the investor at the end. And I say investor, you’re the pensioner collecting a pension at the end of this. You have a need to market at some point. And what it allows her to do is do so much more so much faster. So for the solopreneur, it allows them to do things that they wouldn’t have been able to do, build a tool without having to actually hire a software developer. For her, it allows her to do so much more. Rather than six weeks to get a piece out, she can have it done in six hours.

Sam Carlson (18:45):

I want to clarify something. When you say she doesn’t need content writers, it’s not that she doesn’t need content writers, she needs less of them.

Spencer Burton (18:52):

That’s right.

Sam Carlson (18:53):

Because they’re still creating content. Content didn’t lose its utility. It’s just, “I’m a person. If 75% of the work comes to me complete, now that’s 75% of the work that I don’t have to do. I can look through it, read through it, change whatever needs to be changed, and that can be really fast.” And then coming up with new angles, new ways, that is such a shortcut and an efficiency tool. That’s what I was talking about in terms of Copilot. It’s not replacing people, it’s compounding the skill of the really skillful people.

Michael Belasco (19:34):

And one thing I noticed about self-improvement is that utilizing ChatGPT to help with writing, its vernacular, its sentence structure, it’s incredible. It’s incredibly eloquent, and I’ve noticed that by using it, my own vocabulary writing, all of that has actually improved based on the fact that I’ve used that. And that was an unintentional benefit that I had not thought about until I became so acquainted with ChatGPT.

Spencer Burton (20:08):

That’s a good point. I know far more about Excel VBA after that two-hour exercise, building the first iteration of that Excel add-in than I ever did in the years, and I mean, every one of our models almost has macros just because the back-and-forth conversation with the AI taught me a lot about the file structure, the creation of a ribbon, the syntax of the code, all those things I learned just in that process. And so that’s really interesting. Here’s another use case I want to talk about if I could.

Michael Belasco (20:49):

Yeah, go for it.

Spencer Burton (20:50):

Again, we’re still on the A.CRE, Sam, we might have to do a whole different … Actually, we probably should do a whole different episode about how teams in commercial real estate will, either are or will be using it.

Michael Belasco (21:01):

Let’s keep this as an AI prompt overview and then we’ll transition because I actually … Keep going because I have another thought, but I don’t want to sidetrack, so…

Spencer Burton (21:11):

Yeah, so I think it was Sam, you talked about how it supersizes each one of us. So it’s like putting on a suit that makes us five times more powerful than we were prior.

Michael Belasco (21:27):

Like those exoskeletons-

Spencer Burton (21:28):

Yeah, it’s like an exoskeleton, but for our brain. And so a great example of that is in our Accelerator program we have this Q&A. So that’s where Accelerator members ask the A.CRE team questions. And we have a team, it’s not just Michael and I who respond to these. We have Arturo, we have Emilio who are highly trained in real estate financial modeling, and they provide answers to questions. And then Michael and I answer, call it the more complex questions, and the bottleneck is the human element. And some questions, especially in course one, are fairly elementary questions and no issue with the question. I mean, all of us have to start from zero, and those who are starting from zero, they’re asking these questions and they might have more questions than they feel comfortable asking a team at A.CRE.


And so we are in the process of training, call it a separate version of a natural language model on all of our content in the Accelerator, basically training an AI to know what Michael and I, and Arturo and Emilio know, and not to replace us necessarily, but to give Accelerator members greater access to information instantly rather than having to wait for a response. That’s something we’re doing, and it’s in the early stages of this, being able to train these natural language models on proprietary data like we have. But we’re doing that, and in fact, by the time this season comes out, that tool may be released in our Accelerator AI. And to me, that’s one of the most impactful initiatives that we’ve underway related to AI.

Michael Belasco (23:22):

That’s one of, if not the most exciting thing, I think. Something that can be supplemental … Our core focus is training and teaching people, whether it’s for free, which a vast majority of the stuff is on the website or through the Accelerator program. And to be able to get that out faster and get people learning faster, again, some of these questions can be very challenging to some, but as you’re scaling up the proficiency to modeling, a lot of those could just be programmed and easily answerable. You don’t have to wait for Spencer and I to get out of bed in the morning if you were up late cramming or … So it’s exciting.

Sam Carlson (24:02):

Go ahead.

Michael Belasco (24:04):

Since we’re on the AI sort of primer conversation here, I want to hit on the human element, and I don’t have an answer, I just want to pose the question to the group and have the conversation about it. And maybe we’ll split this episode up because there’s a lot of content here. The human fear, we’ve talked a lot about the optimism and there’s certainly a lot of things to be optimistic about, but there is a real fear, and it’s valid, about the risks involved with bringing and introducing AI, and particularly for people with different personality types, having AI, being a person that wants to train on language and how to prompt machinery or there’s a lot of different opportunities that present itself. And I don’t have all the answers. I’m not going to pretend to know where this is all going, but there are some people, and you see it from the ground.


And I just want to get your guys’ take and thoughts, and I can share some of my own. Graphic designers are a perfect example. And I saw this TikTok the other day, and we were talking about Adobe’s picture completion. And this guy posts on TikTok and he is like, well … and he was being calm and sort of funny about it. He’s like, “I don’t have a job anymore. Watch this.” And it was like a scuba diver next to a coral reef, and he pushes the button and it generates, not only one option, but three options. Something that would’ve taken him days or months. And this is a guy who genuinely enjoys being creative and being in the medium of art and loved it and was fortunate enough to be able to make money off of this. Does that go away? Do these people who thrive in this other type of world get displaced? And not that you guys have the answer, I’m just curious to hear what you think about it. I don’t know what to think about it.

Sam Carlson (25:55):

I think there’s examples of this all over the place. I think you have social media come in, you have cryptocurrency come in, and they evolve what things are. We were talking about your sister who’s running a content division. She’s evolving how she does that.

Michael Belasco (26:12):

She’s always run that though, right?

Sam Carlson (26:15):

She’s always done content. She’s been a content writer for a very long time. I’m an optimist. I don’t think doomsday. I think evolution and change. I think whenever something moves in, whenever there’s a new channel, it’s like a pipeline and that pipeline is delivering whatever is delivering it, and whoever jumps into the pipeline the fastest is going to intercept and take the opportunity. The people who are scared of it and the people who think it’s going to destroy their business, the problem with … What is that phrase?

Michael Belasco (26:50):

Not destroy your business, destroy your job.

Sam Carlson (26:52):

No, but… Okay.

Michael Belasco (26:54):

Let’s talk about the-

Sam Carlson (26:55):

Skate where the puck is going to be. Skate where the puck is going to be. And I think that is the problem with thinking that it’s going to destroy everything. It’s going to change things. And so what is it going to change? In my role doing what I do, what’s it going to change? So in this example you gave, he’s a graphic designer. Maybe he’s not going to be graphic designing images. Maybe he’ll be doing something different at a higher, and doing a value thing. So this will allow him to level up, find the new opportunity. So I don’t think it destroys anything. I think it complements it. What do you think, Spencer?

Spencer Burton (27:34):

Yeah, so I’ve heard this discussion quite a bit over the last few months, and it calls back to these two images that have floated around the internet over the years of, I want to say it’s like 1905 New York City, and it’s almost all horse and buggy. You’ve seen this image. And then there’s a 1915 image and it’s almost all cars. I think there was one horse and buggy in the 1950 image, and there’s one car in the 1905 image. And what that makes me think is imagine your job in 1905 was to manage the horses, and I don’t even know what the word is, but-

Sam Carlson (28:19):

Be a stable boy.

Spencer Burton (28:21):

Well, no, the individual who actually drives the horse and buggy.

Michael Belasco (28:26):

A buggy driver.

Spencer Burton (28:28):

A buggy driver, whatever they called it.

Michael Belasco (28:30):

So sophisticated.

Spencer Burton (28:32):

I have no idea what a buggy driver’s called. But you think about the skills that were needed in order to drive the bug. You have to be able to take care of the horses, you have to understand how to lead the horses. And it was probably a little scary when that first car appears and you go, “My way of doing things is going away.” So what happens to that individual? Well, they learn to use the car and their life actually gets better. They’re not out in the elements. They don’t have to pick up crap.

Sam Carlson (29:10):

Manure I think it’s called, Spencer.

Spencer Burton (29:11):

Yeah. And their life actually got better. And so to me, this is another opportunity in a long string of improvements going back millennia. But the printing press being a great example, when the individual had to hand write every single copy all of a sudden didn’t have to anymore. It elevates humankind. It elevates each one of us to do what matters more.

Sam Carlson (29:41):

So let me jump in here. I’ll give you an example. On the drive over here, we were coming from dinner and I was telling you guys how annoying it is to now Google things. And it’s not to say that search engines go away because you want to be able to research that way too. But there’s sometimes when you want to curate an answer, and for that, I’m going to ChatGPT, put it in, whatever it is. If I’m researching a project, I don’t have to spend all day researching how to do it. I’ll just say I’m considering … And I’ll create a prompt and I’ll get better at making prompts and better and better. And then over time, if I want to put together a marketing initiative, whatever it is, I’m like, okay, well here’s the prompt, bam, “Okay, team, come in. Let’s do this.” It’s just a different way. It’s a curated way. It’s a faster way, and faster isn’t always better. So I don’t want to use that word.


It’s more streamlined and efficient and accurate. It’s changing. Search engines are going to look different. Again, I think it’s a value add to the world. I do not want to be a horse and buggy boy. I like doing what I do. I prefer to be in this…

Michael Belasco (31:00):

You would be so good. Look at that build.

Sam Carlson (31:03):

I mean, I could handle a horse, but anyway…

Michael Belasco (31:07):

All right. So my take on it is that yes, there is fear. My own personal world has, I would say, improved. And there was this old sort of belief that as technology comes, people get more leisure time. Technology was supposed to take over. I don’t find that to be a truth because I think a good majority of people that I know, including myself, are driven to be productive. And so all it does is it opens up a whole new world for people to, again, put on that exoskeleton and be able to do a lot more. To me, in my personal experience with it so far, has been nothing but just an incredible opportunity and door opening experience. But I can see and empathize with the fear of people who’ve dove deep into a career that the writing’s on the wall for.

Sam Carlson (32:09):

Well, let’s do this. Spencer, final thoughts from you on this topic, and then let’s delve into careers and how this is going to affect careers in commercial real estate. So Spencer, final thoughts on AI?

Spencer Burton (32:22):

Yeah. Our mission at A.CRE is to add value to people’s careers, and it starts with real estate financial modeling, but then it gets into careers, it gets into education. This just allows us to add value at scale. You look at the amount of value add content tools and the like that we have been able to put out there in the last three months, and you compare that to the last three years, and that to me is the greatest benefit here. So yeah, I mean, there’s certainly risks and we should be cognizant of those, but I just think it makes things better.

Sam Carlson (33:01):

Awesome. Awesome. Well, this is a fun season for us, you guys. We are here in person and we’ll probably just take a quick break, have another conversation, spin into the next episode. I think I know what it’s going to be about. But thanks for watching. We’ll see you on the next episode.