Speaking at Columbia Business School’s 16th Annual Real Estate Symposium in late 2023, Mark Peter Davis ’08, Managing Partner of Interplay, explored the transformative impact of cutting-edge artificial intelligence advancements on the real estate industry.
Watch a video of his presentation above, and read the transcript below:
Mark Peter Davis: My mission today is to make sure everyone in the room has a foundational understanding of what AI is and what is actually happening in the society around us. I'm probably going to leave you with more questions and answers because we are at the very beginning of a massive journey.
So a quick introduction. I founder managing a partner of a firm called Interplay. Our mission is to architect the future, so that takes a lot of different dimensions. This is what our organization looks like. Just to give you a little sense, everything centers around the singular mission. If you go up, we have a venture capital fund and a multifamily office. We have a studio where we start companies every year and we coach companies and we co-found a number of companies that provide services to startups, and now we also lean into doing things that are explicitly for social good around community and philanthropy.
It's a big team to do this. This team is part of the team and rapidly growing. For those who are interested in startups, which I know this is more of a real estate focus, we do a weekly podcast. My partners and I get on and talk about different, all sorts of different things. It does range from geopolitical to thought leaders, and then there's some startup tactical stuff which may not be as relevant for all of you. And I wrote a book, but I'm the worst book promoter you'll ever meet. You can find it on Amazon if you're trying to raise startup capital, but if you're not raising venture capital, this is a very bad choice. Do not read this book.
Okay, let's baseline a little bit. There's a lot of jargon used around ai. I've actually cut this down to two topics just so we have language and a framework. When you're reading the news to understand what people are saying, ai, everyone can read. This is a simulation of human intelligence and it's machines that are programmed to think like humans. This is limited by the framework and the programming of the technology. It's not creative, okay? AGI, artificial general intelligence. And you should pay attention when you read the news and you see that extra word in there can perform tasks basically as well or better than you and can adapt and learn and be creative. And this is what we really all think about when we watch movies. A GI, you're not thinking about ai, but the only thing we're using today is ai. So the good news is a GI does not exist yet. The reason people are scared is because it might exist very soon.
So let me give you a way I think about the significance of technology of this technology. AI in particular, I think it belongs on this chart. When you think about the milestones that have defined mankind, it's a couple of dozen things. I think it's on the sheet. So we are at the very beginning of a major moment that people will look back on in history books and say, that's the date when that happened. Now the reality is that curve looks a hell of a lot more like this than the curve you just saw. Modern man is estimated beyond this planet about 80,000 years, but all of the innovation that we count as relevant happened the last couple of hundred years. When we think about our history, we think about a very myopic period of time. Reality is it's all changing very fast in ways we're not ready for, and frankly most of us are not biologically adapted to.
Okay, so what's the big deal about ai? Why does the tech community care about it? To have an understanding of that, you have to have a little context. So in 2000, we have the internet and market. We're in the tech boom and we're producing a lot of data. Anyone who wants to get on and write a blog or anything else is doing that. Emails are flying around the internet. In the last 20 years, we've thousand x-d that it's a huge amount of change. So the amount of data that is available on the planet earth far exceeds any other time that you can imagine, and it's way more data than we can use. What we are doing today in all of our business decision making, geopolitical decision making, your political decision making is based on simple small data sets that you can read in a newsfeed or consume from friends. The reality is every transaction, every inch of the global map, it's been documented, it's been considered. You just can't process it. You're not smart enough, and that's where AI comes in and why it's a big deal. Now we're in the very beginning when you think about that, take a step back is you can think about AI as the first brain that can actually make sense of the data that we now have. That's the missing piece.
We're in the very early days of this journey. So chat, GPT, only the really launched a year ago. I feel like it's common vernacular for most people in the room or read the news, but the reality is most people haven't even heard of it or not most, but a lot of people haven't even heard of it yet, and even fewer have used it. Who in the room here is logged on and asked questions? Okay, good. Okay, you're in the minority. What's that? Students have, students have hyper science. There you go. There's a fancy business school curve here, which I'm not going to get into for the phases of innovation. Suffice it to save from this diagram. AI is in the very, very, very beginning of its curve and the data, it's processing is going way up. Just in the course of the last three years, the amount of data ingested and being leveraged by these models is five x-d and things tend to not change linearly. Look at the slope there.
Here's a way that we can all make sense of to appreciate how much more powerful this technology has become in just a short period of time. Chat, GPT-3 0.5 scored a 1260 or so in the SAT six months later 1400, okay? How many of you lose to chat GPT now? Okay, here's how I think about the opportunity, and this will help frame for those who are building frontier. This is a little bit of our investment thesis around where AI fits into the world. If you look at cloud computing, there's kind of two buckets. This is an oversimplification. Any tech nerd would scream at me, but that helps you see the signal within the noise. There are platforms and applications. The platforms are the heavy tech that everything else is built on. And so if you look at cloud compute, you've got Amazon web services, Google, Microsoft, they all have products as well.
Those are hugely capital intensive, hard to build, and then everyone else builds tech on top of that. And so almost every application today is built on a cloud computing backend. Okay? Same paradigm in mobile software. Reality is there's two platforms. You've got Apple's App store, you've got Android, maybe a couple others more or less. It's pretty concentrated, huge development task. Most startups, the way the institution that I'm in, in venture capital is set up, they're not set up to fund those types of scale of companies, so they're owned by big tech and then everyone else enters and builds application layer technologies, and that's the stuff we all think about and use. Here's the fun thing to think about. It's a real long tail game. It's just a lot of very little of what's built actually as relevant for our lives and what generates real yield.
So just to give you a little context, two million-ish apps on the Apple App store. It's really a hundred apps that matter. That's where all the value is created economically for investors. And those are the same a hundred apps on almost everybody's phone in here. We think that's the framework to think about AI and that should say ai, not mobile software for clarification. So don't get confused. You've got a handful of platform companies that are investing hugely in developing extremely capital intensive technologies, right? We're all talking about open ai. There are a handful of others.
My background intel is that a lot of the big tech companies had AI in development long before OpenAI showed up on the market, but their market entry catalyzed the need for those big tech companies to start exposing their underlying software. So they already had it and were using it. It was just not publicly accessible. There's really little in the way of application layer that we're currently using. Maybe you're logging into chat GPT to ask a question. That's it. It hasn't shown up in any material mobile app that you're using, any application in your computer anywhere that's affecting your life yet. These are the earliest of days. We're all talking about the infrastructure, the platform layer still, we're not really talking about relevant applications, but they're at the gate. There's a lot going on and it's going to penetrate your work life. Just a little quick glance here, I'm not going to run through this, but there's a lot of companies the impression I want to give you here already targeting FinTech a bunch thinking about healthcare and a hell of a lot thinking about content to understand where this is going, let's take a second and talk about convergence.
Because AI is a ubiquitous platform and the most interesting applications of it are not going to be what we would consider AI specific. They're going to be changing the way we do other things that we're already engaged with. It's an undercurrent that's going to affect a whole lot of industries, an easy one that I think everyone can see because everyone's watched a lot of Marvel movies, AI plus ar. If you imagine Tony Stark's visor in the Ironman suit, that's what we're talking about, an intelligent visual overlay on what you see in the real world, and there's a lot of real applications coming with this. They're not market viable yet. We don't have the hardware in place, but this is coming. And so imagine anytime you need to make a decision if the world was color coded, if you're pop in the hood of a car and you've never fixed an engine in a car before and there's an instruction manual integrated into your line of sight, or we found this one because we thought it was kind of real estate-y. If someone's carving up one of those model things, here's a digital one. I don’t know if that's actually going to happen.
Okay, here's another one, AI plus robotics. Who in the room has heard of Andril? They were in the news last couple of weeks. Okay, that is scary. Just of the couple hundred folks here, maybe 10 hands went up. The founder of the Oculus gentleman named Palmer Luckey has started another venture. It's already worth billions of dollars. And what he is building are drones that can fight in wars on land, air, and sea. Real drones, land, air, and sea. There was a big press announcement this week. They have a drone that can fly up into the air smack into another drone, and if it doesn't fly back and land, the nature of warfare is a hundred percent changing. And when you mix that with ai, it's the background for a plot of a movie we've all seen too many times. Doesn't mean it's going to happen, but these are the ingredients for this story. Alright, who's seen Tesla's in-house robot? Okay, if you haven't worth a minute, go to YouTube at some point. Look on Tesla robot and you will see something uncanny. There will be a humanoid looking robot standing in a room sorting blocks intelligently, and some human will come up and scramble the colored blocks around and the robot will figure out that they've moved and sort them appropriately. Two or three minutes of your life that's coming, we will have robots in our homes. There's another movie plot.
All these things are happening now. These are no longer sci-Fi. These are in development. Okay, here's another one, AI plus quantum computing. All of us have heard the phrase quantum computing. It's a daunting physics-y sounding word. I assume no one understands it, really. I probably don't, but here's what you need to know. This is basic English around it. Today's computers are 10,000 times faster than computers 20 years ago. Huge progress. Quantum computers are expected to be 158 million times faster than the computer you own today. And they're in development. They do not exist yet. They're in development. To put that in context, what will take your computer 10,000 years to do? This computer can do in 200 seconds. If you're not scared yet, keep listening.
This is Jack Hidary. He's also Columbia grad. You don't know his name. You should. He has been a very successful entrepreneur in New York, I believe he ran for mayor of New York at one point. He took two companies public and then joined up with Google and he spun a company out of Google, and I think Google has only spun a handful of companies out. And that company is called SandboxAQ. And they were seated with a $500 million check to connect AI and quantum computing and make it real. And this is a big deal for technology. It's a big deal for national security. It's a big deal for everything that you guys are thinking about or worried about or losing sleep about, because when this technology is out, every form of cybersecurity currently in market today is hackable instantly. Everything you've password protected is visible to someone with this technology. Every document at the Pentagon resume is visible and the world changes.
And their first client, the Pentagon is a good one. I think it's likely that drug discovery won't require clinical trials. There's something called synthetic data. When you can run models powered by supercomputers enough, you can create enough data to actually figure out probabilistic outcomes without actually doing experiments. And so it should accelerate all types of medical benefits. That's a good side of it. The good news is these computers don't exist yet. The problem we have is we've outsourced all of our manufacturing, more or less in the states for our policies. Manufacturing of semiconductors and chips is so capital intensive that it's too expensive for a single country to fund more or less a real program. So basically the whole world is aggregating all their demand to a couple of locations, the hub of which is Taiwan.
Taiwan produces 60% of the semiconductors, but what matters is they produce 90% of the advanced ones and to make matters more fun, it'll take the US 10 plus years to build an equivalent technology set on our ground. And policies are being put in place and things are happening. But you can see a lot of these geopolitical conversations have a very powerful undercurrent because we're talking about whoever gets AI plus quantum computing first gets to see everyone else's stuff. Okay, so that's a little primer on ai. I'm going to spend a couple minutes on real estate and then answer questions. They brought me here to talk about this and I was like, I don't know anything about real estate, so I'll just do this quick. This is actually one of our portfolio companies, a company called OnsiteIQ. They have folks walk through properties around the country wearing these devices. They take 360 video images of the space, AI then weave the video together to create a viewable floor plan that someone can actually navigate, and then it intelligently identifies changes so that the idea is a lender or a developer who's building a building across the country or across the world and monitor progress remotely. So for all of you who are gas guzzling with private jets, maybe you don't have to do that anymore. Check out on Site iq.
There's a bunch of these, but just let me give you four quick examples. Realtime AI video analytics. So maybe in properties you'll be able to identify threats like someone's carrying a gun into your building. They're painting walls now with AI and computers, so no humans required. There's tons of ways to communicate customer service. This is the area that the barbarians are hitting. First customer service jobs through chat and intelligent ai, identifying, using cameras, unforeseen risks that can cause human harm. I share with these, give you these just to give you a few anecdotes of how to think of ai. It's going to affect everything that you guys are touching as the headline, not one piece. And so the questions I'll leave you with is, are you minimizing costs accordingly? Are there tools out there that can help you in your version of the real estate market? Are you maximizing quality? I don't believe all of you are incentivized to do that, but wish you would.
How will you invest differently? And I don't just mean which properties. How will your investment decision-making process change right now? How many variables do you take into consideration when you determine the highest and best use of a property or what region you're going to operate in? It's whatever you can make sense of or stuff in an Excel sheet. What are there something a little bit more intelligent helping you and how will you hire differently using Excel is becoming an unnecessary skillset. There's AI now where you can talk to the Excel and it builds the charts and diagrams. You just need the data and the intelligence to talk to it. So the shape of teams and skills required from institutions like this is going to change. Last thought I'll leave you with is I think the history of technology is riddled with a phrase I use as the butter knife. To me, a butter knife is a tool that a mankind is made that can be used to spread butter or kill somebody. And it all depends on what the holder chooses to use it for. AI is for sure a butter knife. The question is how will we use it?