Raise Capital Smarter with AI, with William Hollis and Donna Mitchell

Donna Mitchell [00:00:00]:
Welcome to Pivoting to Web three. And I've got a real gem here today because he is from my home area, New York City, and he comes telling me that there's a certain restaurant still on the corner. I'm gonna go and say it. White Castle. I used to love me some White Castle, and it's still on the court. I gotta say it. So anyways, let me get back on track. I had to let you know who I'm talking to, but his name is William Hollis.

Donna Mitchell [00:00:24]:
And William is automating, raising money, scaling, doing everything he can do in real estate and bridging the G gap between the technology and raising capital. And I haven't had that on pivoting the web3 yet. So I think he's going to be somebody that we want to listen to, talk to, and probably connect with. So, William, say hello to your audience in ours and just tell us how you ended up where you are today.

William Hollis [00:00:51]:
Absolutely, Absolutely. Thank you so much for having me on, Donna. Happy to talk to a fellow New Yorker. For sure. Not too many people know about the White Castle. No. Not to be.

Donna Mitchell [00:01:01]:
Let it out the bag.

William Hollis [00:01:04]:
I love it. I love it. Yeah. So William Hollis. I go by Hollis. I grew up in Hollis, Queens, believe it or not, a few blocks from Hollis Avenue. So the name kind of stuck. You know, I started out, I've had a lot of different jobs in my life, but how I landed on being entrepreneur, my dad kind of put this entrepreneurial bone in our.

William Hollis [00:01:25]:
In our body. When I probably seven or eight years ago, I started looking into really real estate investing. And I tried a few different things, did a few deals, but really what I wanted to do, what I wanted was the quote, unquote, passive income that everyone's after. Right. We hear that term thrown around a lot. And so I was introduced to real estate syndications, which is basically, if you think you see a giant apartment building, you know, in your neighborhood, one person doesn't own that building. A bunch of people own that building together, and they kind of pool their money together to go out and buy those large assets. And for a lot of those investors that have pooled that money, that is really passive income.

William Hollis [00:02:05]:
They don't do anything. They don't do with any. Any of the three T's, the tenants, toilets and trash. They don't have to deal with any of that stuff. Right? So I said, okay, that's what I want to do. So I got started to get into this space and realized, if you want to buy a big building, you're Going to need some big money. So I started to learn how to raise capital. All the while I had this AI and software background.

William Hollis [00:02:27]:
And so when I started to raise capital from my own real estate opportunities, I realized that there was very little technology being used in the field. Especially when I got started, slightly better now, but especially back then, people had actual Rolodexes of numbers that they were called to try to raise equity from. So I said, let me try to bring you guys into the 21st century and see what we can do. So that's kind of the buzzly organic growth of my company. It was just solutions I built for myself and then I was able to deploy them for others in the space. And I was doing it, believe it or not, for free for a very long time, just because I love technology, I'm a nerd at heart, and I like to build cool stuff. And then one day someone said, hey, would you do this faster if I paid you for it? And I said, I probably would. So thus, you know, raise AI was born.

William Hollis [00:03:23]:
That's our company now. We use AI and automation to help raise capital for private equity investment opportunities.

Donna Mitchell [00:03:31]:
So thank you so much for that explanation. So for those that are listening that are in real estate, and that's quite a few folks out there today, either they're flipping, wholesaling or counting doors, and they're in that space. You have AI in the market. I've had a couple of guests talk about AI and urban urbanization development. How specifically should we be looking at AI in the real estate market? How does that improve? Is it in the search? Is it in aligning of the type of property you're looking at, or give us a little bit more? Could you paint the picture or use case and how AI and real estate come together?

William Hollis [00:04:12]:
Absolutely. So I'll take you really. AI can be involved in as many areas that you're, you know, you're willing to plug it into. I'll give you two or three different use cases here on the acquisition side. Right. One of our clients underwrites and, you know, by anyway, I mean like kind of doing an analysis of roughly 150 properties per week previously. So that's him finding these opportunities, whether they're sent to him by a broker or, you know, he's looking on the Internet on his own, and he's taking those numbers, those properties, plugging the numbers into his spreadsheet and analyzing those deals to see if he wants to, you know, move forward or make an offer or not. Now leveraging AI, he's analyzing 150 properties per day.

William Hollis [00:04:59]:
Oh. So as you can imagine, you know that's just taking weekdays, right? That's 5x. His ability to analyze deals and to get eyes on opportunities. Now, taking that a level, a level deeper, AI has the ability obviously to pull data, real time from data sources. So now your underwriting isn't just estimation. You can get real time estimates for things like repair costs for, you know, different debt products that you might need to use. You can get real time estimates so you have more underwriting that's able to be done and more accurately. Take it a step further.

William Hollis [00:05:36]:
The AI will also is also able to reach out to the property owner or real estate agent or broker who sent you the opportunity with, with a letter of intent for purchase of that property. So if you think about it, zoom out a little bit, you say, okay, we underwrote a thousand properties this week and the AI knows my buy, my buy box, we call it, right. My criteria for purchase. It found those properties, underwrote them with real time insights, then it sent letters of intent to all the good ones. By the end of the week, let's say I sent out out of 100 properties, probably sent out seven offers. Realistically, out of a thousand properties, sent out seven offers and now you have a response back on three of them. Wow, are ready to buy three, potentially buy three properties. And you didn't actually have to intervene as a person.

William Hollis [00:06:32]:
Right. So that's on the acquisition side, on the capital raising side, we're able to do something similar. We're able to train the AI on what types of investors would be interested in the investment that you're offering in your real estate or whatever kind of private equity investment you may have. We're able to train it on who would be interested in that. Now we're able to use that data to put together ads or marketing campaigns to reach out to those people. Specifically when those people engage with those marketing campaigns. The AI is able to act as kind of a junior level salesperson, you could call it, to engage that person, answer any preliminary questions they have and have them schedule a call with, typically with, with an investor relations agent for one of the private equity companies. So they that that call is scheduled, that lead came in, was educated and nurtured before a human even had to had to intervene.

William Hollis [00:07:28]:
So that's increasing the number of leads that are coming in and the speed to closing that capital as well. And there's a bunch more things, but I won't hog all the airtime here.

Donna Mitchell [00:07:39]:
No, you can hog the Airtime. Because right now we're learning, okay? Trust me, we didn't miss the beat. So with all the speed and efficiency and effectiveness, it really does save time and maximize opportunities and minimizing costs. So those that have a fear of being replaced, I always felt the best way to look at it is you get enhanced with enhanced results in your process or your workflow or who you are. So with that said, and you've had a lot of success with that success, are there any areas that you've seen that concern you or you an area where you wish there was more development or people paid more attention to it? I guess. What do you raise your eyebrows at.

William Hollis [00:08:27]:
Today in the world of AI? All of it. Level set. I've been working in this industry. I like to say, before it was cool. So 2013, I got my. I got a job at what was at one time the number one or one of the top, I'll say one of the top AI solution companies. Enterprise level AI solution company. So we consulted for JPMorgan Chase, E Trade, PNC, building AI and automated solutions at that level.

William Hollis [00:09:02]:
Back then, AI was not all that intelligent. It was good, but it was not that good. Um, now AI is very good. Very, very good. There were originally a few rules in the AI community that we said we wouldn't break. I'll give you two of them. Don't give it access to the Internet and don't teach it how to code. And we did both of those.

Donna Mitchell [00:09:27]:
Both of those.

William Hollis [00:09:31]:
Now, when I say concerned is because. And this goes with any piece of technology, Right. Anyone with malicious intent will find a way to exploit the technology. And AI has made technology very. Technology in general, very exploitable. I'll tell you where I'm not concerned, conversely, is with the loss of employment. Because every technology boom removes opportunities and creates new ones. So there's a stat that.

William Hollis [00:10:07]:
I might be misquoting it here, but the stat that in 1914, 60% of the jobs that exist today did not exist back in 1914. Right. So I'm sure the flip side of that is 60% of the jobs that existed back then don't exist anymore. Right. But. And that's just because technology grew. My mother is a great example of that. My mother was secretary for the CEO of bellsouth.

William Hollis [00:10:33]:
She was the type of secretary that used to go in with her pen and pad and, you know, write correspondence from the CEO just by hand. And then the typewriter came along. And I'm so sure all the secretaries thought, oh, no, we're going to be.

Donna Mitchell [00:10:47]:
What happened to my shorthand?

William Hollis [00:10:50]:
So my mother taught me shorthand. That's a perfect reference.

Donna Mitchell [00:10:53]:
Hey, I'm there, I'm there. Okay, I got it. Just like the White Castle.

William Hollis [00:11:00]:
But I mean, yes. Are secretaries really a thing? Not in the way that they used to be, but those people simply had to learn other skills and go do other things with the new technology that was available. So those are kind of some of my fears. But at the same time, I think there's a lot of positives, like I said, especially when it comes to the fear that a lot of people have of being replaced by AI.

Donna Mitchell [00:11:26]:
So if it's okay with you, I want to go back and talk about why didn't you want it to have access at the Internet, even though that's what we ended up doing? Because it's going around the Internet now as we speak. But what was the concern that that was one of the two things you weren't going to do and did it anyway?

William Hollis [00:11:44]:
Yeah, because we always do that. Right. That's. That's human beings for you. That's human beings for you. Well, one AI currently has the ability to present misinformation and to propagate misinformation in a more prolific way than ever before. So let's just say, for example, I wanted to attempt to manipulate an election, right? Because my AI has access to the Internet, I can have it go out and find specifically the kind of messaging that will sway a certain group of people in a certain direction, because AI has not only access to the Internet, but the ability to code. And now there's AI virtual machines.

William Hollis [00:12:39]:
I can have it go create 5 million Facebook users that will push this agenda that I have in a very human like way. It won't seem like it's a machine saying these things to you, but it will be a machine. More machines than we could stop. And so in a large way, when AI has access to the Internet and the ability to code, technically it has the ability to go off and create things unsupervised. Right. Having the ability to code means that AI and AI could set up a website and do business by itself if I program it to.

Donna Mitchell [00:13:26]:
That's.

William Hollis [00:13:28]:
That opens up some serious security risks and vulnerabilities. Not only maliciously, though. I could unwittingly create something malicious. But I won't scare the people. It's.

Donna Mitchell [00:13:46]:
So we don't want to scare them. Okay, so. So we're gonna. We're gonna be careful on. On what we say. But I've been wanting to ask somebody this question or, or delve into this as a. I would say let's delve in. My question is, do you think society has been engineered to be where we are today? I'm going to use that term, engineered.

William Hollis [00:14:17]:
I'm not maybe the best person to speak on it, but from what I can tell, absolutely.

Donna Mitchell [00:14:25]:
Okay, we'll leave it there. I've always just kind of thought about it from a business process standpoint or culture change standpoint. And when I look at where we are and where we've come from, I just always wondered, is there a possibility that we've been engineered over the last 10, 15, 20 years? Why we turned the way we are.

William Hollis [00:14:50]:
Since the Second World War, really since the first. But the Second World War and the boom of. I'll take it back, I'm going to say the First World War because that triggered a boom of marketing and technological growth that was just not seen before.

Donna Mitchell [00:15:15]:
Right.

William Hollis [00:15:15]:
And those are the two things you need to engineer a society. You need propaganda and technology that can propagate that propaganda in a very prolific way.

Donna Mitchell [00:15:28]:
Right. So let's get back to Web3 and AI. So we don't want to turn this into a conversation that may not be able to maneuver between the cracks too well. So at the end of the day, so here we are, we're in real estate and you know, you have blockchain and blockchain has smart contracts coming. Have you done anything with smart contracts yet? Have you heard about that?

William Hollis [00:15:49]:
I have heard about it. I have not done anything with it yet.

Donna Mitchell [00:15:54]:
So I'm curious to know. We won't go the blockchain direction. Let's go back into the AI space in real estate. What biases have you seen in AI and can we fix them, can we correct them, or have they improved since AI has come to market, especially in the real estate space? What do we need to be concerned about or what can be modified?

William Hollis [00:16:18]:
For sure, for sure. So I'll talk about two different lanes here, because when you say bias, I think kind of human bias toward AI, but then also AI itself can have biases. So with human bias towards AI, I think really it comes down to our fear of change. I think AI is, is pretty. It's a pretty scary thing. Just in general, even on the good it can accomplish and how fast it can do it. Right. That's a pretty scary thing.

William Hollis [00:16:48]:
I think people are fairly resistant, especially in the real estate space where you're talking about such large sums of money. Right? So typically in our, our client base for raise AI, it's not unusual for one of our clients to have an investor who's sending over 200, 300, $500,000. When you're dealing with such large sums of money, you want to make sure you have an eye in all the details and really be able to touch any. Everything that's going on. And AI using AI, some of that, right? Because you're relying on this intelligence to handle certain interactions or to know when to do, when to do certain things. That being said, at the end of the day, and we, we see it in our, within our client pool, those who are more willing to adopt the technology will simply move faster and be able to accomplish more, especially as time goes on. And it's a matter of competition, right? It's kind of like you, you're in this NASCAR race and some people are deciding to use this component that helps them go a little faster. And some people are saying, no, I'm a better driver, I can do this by myself.

William Hollis [00:18:02]:
And eventually the people with that little component are gonna, you know, they're gonna win. At the end of the day, they're gonna win. So that's the viewpoint I encourage everyone to adopt when it comes to AI, is like, listen, this is here. If I don't begin to at least begin to understand how it works, in very little time, and by very little, I mean five years time, you'll have no chance. You'll just. Everyone will be running circles around you. So you don't need to be an early adopter, but be an adopter. Be an adopter.

Donna Mitchell [00:18:39]:
So when you contacted pivoting to Web3, what message was it that you really wanted to share, that you wanted to get out there in regards to what you're doing or the projects you're involved with, or the vision that you see with real estate and AI?

William Hollis [00:18:52]:
Yeah, I appreciate you bringing that up because really, kind of the gospel that I preach to people is to lean in. Don't be afraid of it. Lean in. And you brought up a point. I can't remember if it was before the show or not, but it's a matter of letting AI enhance your natural skills and abilities. So that actually brings us kind of full circle into the biases that AI might have, because this is something that I really want people to understand. You get out of an AI what you put in. At the end of the day, you get out of an AI what you put in.

William Hollis [00:19:34]:
The best thing that any of us can do, whether you're entrepreneur, investor, executive, whatever position you might have or whatever career you might have is learn how to speak to AI effectively. They call it prompting. Right? But that's just a fancy word for how you speak to AI. And learning how to speak to AI effectively might become one of the single most valuable skills that we can have over the next 10 years. Because that is how you, how you maximize the results, how you maximize the intelligence the AI uses in your behalf is by knowing how to ask it the question in the first place.

Donna Mitchell [00:20:26]:
So for those that are listening, let me share this. The conversation that I have with AI now is not the same conversation I had last year. The conversation has improved. What I put into, as you're saying, William, how you communicate with AI. I had a guest come on the CEO of Delish and she said to me, AI became her best friend, her friend, her boyfriend, her girlfriend, and everybody is all right there. So that brings me to this question because I see it now. My questions started being nice. Thank you, please.

Donna Mitchell [00:21:03]:
Oh, that was great. Thank you. You're wonderful. Please continue. Would you like me to do this? This, this? I really would like it all. Is that okay? Talk to it like you got a person there. Be really polite, you know, fluff it up when you need to. Edification to help, but be specific as if you were talking to your best friend or an employee that you want to keep for a long time around.

Donna Mitchell [00:21:33]:
So with that said, how did reasoning come in? How did the reasoning start developing in AI? I've always wanted to ask somebody that question. And since you've been engineering and you've been in all the solutions, you seem like the great, the best guy to ask right about now. Like I say, the great one to ask. That's what I was about to say. So let me just give you some edification on that one. How did reasoning come in here? Because it seemed like I'm talking to a real person.

William Hollis [00:21:56]:
It's. It's incredible. It's incredible. The models that are available now. I kid, I tell you not, I kid you not. When, when I first started, the big thing that AI was able to do was natural language processing, nlp. This was the first job I had. Entry level position was improving the AI's natural language processing.

William Hollis [00:22:21]:
But back then, the AI was really a glorified search tool. Really at the heart of it, when you pulled back all the layers, it was really just glorified search because it didn't really have the ability to make connections or form new connections between disparate pieces of data. Right. This is still many years ago, but after, I want to say 2013ish. At least that's when I started to hear about it. That's kind of when OpenAI was in its infancy. These neural nets became a thing. And to break it down, a database is very structured like line by line, and all the information is processed in a very structured way.

William Hollis [00:23:08]:
If you want to make a connection, you need to have a very specific path between this database and that database and this information and that information. Neural nets are the opposite of that. Neural nets are like here's this data kind of, if you think about it, in a cloud form. And the AI is able to make connections between pieces of data based on the things contained within them. And now that was like the infancy of the, of these reasoning models, right? So now these models are able to use essentially prediction for simple terms, but to make these connections and to understand at a deeper level what you're, what it is that you're asking for.

Donna Mitchell [00:23:56]:
And that's the, is that, is that, I'm sorry, is that the neural networks?

William Hollis [00:24:00]:
Yeah. Yep.

Donna Mitchell [00:24:01]:
Okay, yeah.

William Hollis [00:24:03]:
Or, and that was the kind of the foundation of it anyway. And as it's making, able to, it's able to make these connections and the beauty of it really is the amount of data that it's able to analyze that increases the likelihood that it's able to understand what it is you're talking about. So AI that we use right now is still very much probability based, but the more data it's able to consume, the better it is at making sure that it lands on the right side of that probability of knowing what it is that you need. And now it can compare your need to millions and millions of other needs of other data points and say, okay, if she asked me, for example, if she asked me how long should I cook chicken, a chicken breast, she probably needs some recipes as well, because last time that was, the next thing somebody asked me was for a recipe. And if she needs recipes, maybe she'll need a grocery list, maybe she needs me to buy the groceries. But it's had so many repetitions, you know what I mean? It has so much practice and that's really the key behind the reasoning.

Donna Mitchell [00:25:16]:
Okay, so from the reasoning you now have, the personalization taking place in AI, especially on the retail side, is very beneficial for someone to know that this is your customer that's been very frequent. And when they come in, they buy certain items. And if they're buying these items, they might want these pair of pants. And if they buy these pair of pants, this is the type pants they like. And this is the Designer they wear and these are the colors they wear. So if, why don't we go ahead and get that guy at a tie that he didn't look at yet but he's gonna like. Is that how that all kind of developed?

William Hollis [00:25:50]:
Exactly, that's exactly that. And that's exactly why ads are so effective. And also how you can engineer a society.

Donna Mitchell [00:26:03]:
We all went all the way around.

William Hollis [00:26:04]:
Okay, but yeah, no, to your point, that's exactly how it works. It's analyzing massive data sets to identify patterns where before we had no real ability to do that. If you did it, it'd take forever or you'd find one or two insights. But now there's enough compute out there to analyze incredible amounts of data.

Donna Mitchell [00:26:27]:
Okay, so even though I didn't want to get into this, but you brought it back a second time and we then came back to engineering society. That's a good conversation. I know that's a good conversation, but I don't know if it's for this. But okay, so here you have society who's gone through a culture change and we've got AI in the mix and everybody's getting their news from wherever they want to get their news because they like the support and the alliance and the connection they feel with those like minded people. So everybody's in their own silos and here comes AI la la la la la down the street coming into the culture. Not only could you have AI, but those that really wanted to manipulate everything anyway, utilize the AI in a manner that magnify some things, diminish some things, and then start using emotion and content.

William Hollis [00:27:33]:
Okay, so I'll say this, I heard this from another creator. I believe she's from New York as well. I can't remember her, her full handle on Instagram. It's Nikki something. She had this phrase that says, whoever has the content has the credibility. That's the society that we live in previously is whoever had the experience had the credibility, right? Whoever had the skill, whoever had the track record had the credibility. We no longer live in that society. In the age of social media.

William Hollis [00:28:04]:
This is nothing to do with AI. It's social media captures so much attention, right? And it's so good at feeding you micro bits of news and micro bits of information. And the algorithms, even before AI came into them, the algorithms are so dialed into what we pay attention to that whoever has the ability to proliferate content has the credibility, they have the attention, they have the eyeballs, right? So it's pretty interesting. Time to be alive.

Donna Mitchell [00:28:40]:
So, so, so now let's circle on back to the real estate and raise AI. So you have real, you have raise AI in the real estate and the ability to elevate and raise capital. So if anybody got any sense today, what do you want to tell them so they can do what they need to do with the AI and the real estate and get some capital and get moving in life.

William Hollis [00:29:07]:
Absolutely. So one of the biggest fears I want to address is that people believe AI will mishandle. Well, two things, actually. People believe, one, AI will mishandle investor relationships. And they believe that the investors don't want to talk to an AI, Right. So they think the AI is going to fumble the relationship, which it can if it's not properly prompted. Which is why I said the first thing is learn how to talk to the AI so you can teach it to interact in the way that you want it to. But second, they have this, we have this belief that, oh, if I send a text message or there's a back and forth text exchange with Donna, and Donna's not going to like that I'm using an AI.

William Hollis [00:29:55]:
Research has shown that 40% of people would prefer to speak to an AI. Prefer, not to mention whatever percentage don't care either way, as long as it does whatever it is that they set out to do, which is probably the other 60 is like, whatever, bro, like.

Donna Mitchell [00:30:13]:
Just do, just give me the transaction, okay?

William Hollis [00:30:18]:
Just do the thing. Who cares? Right? So that's a, that's a. Again, that's probably born of an internal limiting belief rather than someone coming up to them and saying, oh, you're using AI, don't ever do that again. It's probably born of. So like more of an internal, an internal fear. However, however, you have to know when to use AI and when not to use AI. And specifically I'm talking about conversational AI or chat AI, which is a huge thing that we use in our. On the investigation side, there are certain circumstances when you should just talk, have it talk to a person.

William Hollis [00:30:56]:
Just go talk to a person. Don't try to put AI in the mix. Those circumstances are anything that triggers a deep emotional response in the end user. I'll give you an example. In real estate, if an investor messages in and says, hey, I did not receive my distribution for this quarter. That's not a moment for AI to back and forth and ask questions and investigate. No, okay, this person was expecting money and they didn't get the money. Let's just have them talk to a person who can de.

William Hollis [00:31:30]:
Escalate the situation. And guarantee a positive result. The time savings of the 15 minute conversation is not worth an irate person. Right. So we would see this all the time where people would, you know, back in the day, message in and say, oh, my phone is not working. And the AI bot would be like, oh, okay, well, what's the issue? Can I do this or can I do that? It's like, bro, my phone doesn't work. Can I just fix it now? I don't want to have this bot interaction. I need a human who's going to help me resolve my problem immediately.

William Hollis [00:32:03]:
So while I do believe in AI, obviously, like we recommend everyone use it. Also understand when it's time to reel it in, program it to stop, like, know when to shut up and just hand the conversation to a real person.

Donna Mitchell [00:32:22]:
Okay, so that's good to know. And that has happened probably to a lot of people too. So how can people reach out to you or chat with you? Who's your customer? Who you. Who do you work with?

William Hollis [00:32:37]:
Real estate private equity companies that have more than 10 million under management are our primary client. Who we serve, we serve a lot of people in adjacent real estate verticals as well. Lenders who are looking to find, you know, more borrowers, real estate community leaders. If you have a community, real estate investment community, and you're looking to bring people into that community, those sorts of things. I'm very active on LinkedIn. I try to respond to every message that's not a sale, a pitch. So definitely shoot me a Message on, on LinkedIn. I'll be happy to connect.

Donna Mitchell [00:33:15]:
Okay, so thank you so much for being with us and I've learned a lot. I'm sure our audience has learned a lot and I appreciate the head up in certain areas. And thank you for talking to us. This is William Hollis and Donna Mitchell on pivoting to Web three and we're shaping the MAR together.