Zach Aarons discusses AI's impact on real estate, energy bottlenecks, and MetaProp's investment strategy
Summary
Zach Aarons of MetaProp explains how AI is shifting from cloud migration to application building in real estate, construction, and insurance. He argues that power availability, not chip supply, is the primary bottleneck for AI growth, predicting solar dominance due to cost and scalability. Aarons details MetaProp's barbell investment strategy, avoiding competitive Series A rounds in favor of seed and late-stage deals. He also discusses the challenges of AI maintenance for incumbents and the potential for adaptive reuse of office buildings into data centers.
Episode sections
- 0:23 – 12:07
Metaprop's Investment Thesis and AI in Real Estate
Zach Aarons introduces Metaprop's focus on AI applications for the built world, noting a shift from cloud migration to building AI on existing data for efficiency gains.
- 12:07 – 17:53
Incumbents vs. Startups and AI Maintenance Challenges
Aarons argues incumbents lack internal teams to maintain 'vibe coded' software, preferring established platforms. He notes a mid-2026 trend of CFOs curbing token spend.
- 17:53 – 23:00
Robotics in Construction and Data Collection
Discussion on teleoperated construction robotics and Figure's strategy of paying users for household task data to train world models.
- 23:00 – 26:51
Energy Constraints and Solar Dominance
Aarons identifies power, not chips, as the primary AI bottleneck, predicting solar dominance due to cost, scalability, and lack of political opposition.
- 26:51 – 31:35
Mortgage Underwriting and Housing Affordability
Aarons highlights AI underwriting tools like Lens AI and material science innovations like Branch Technologies to reduce construction costs.
- 31:35 – 33:45
Venture Cycles and Liquidity
Aarons suggests venture fund cycles may not compress as companies stay private longer, but secondary markets offer new liquidity.
- 33:45 – 35:14
AI in Insurance and Accounting
Speaker_2 identifies insurance and accounting as 'triple bullseyes' for AI, noting that controlling distribution protects against disruption.
- 35:14 – 36:55
Investment Strategy and Barbell Approach
Speaker_2 explains MetaProp's barbell strategy of entering early (seed/pre-seed) and late (Series C), avoiding competitive Series A rounds.
- 36:55 – 39:05
Office Real Estate Rebound Prediction
Speaker_2 predicts a rebound in office real estate within five years due to AI-enabled adaptive reuse into data centers and warehouses.
- 39:05 – 40:20
Career Advice and Mentorship
Speaker_2 advises against 'scrubbing floors' in interviews, recommending candidates show a unique perspective instead.
- 40:20 – 42:05
Dormroom Fund and College Investing
Speaker_2 recommends Molly Fowler's Dormroom Fund model for building a community of college student founders.
- 42:05 – 43:46
Tennis and Networking
Speaker_2 cites tennis as a networking 'secret weapon' and mentions Zach Schweida and Michael Chang.
- 43:46 – 46:11
Japanese IPO and Community Events
Speaker_2 notes an upcoming Japanese IPO and promotes MetaProp's podcast and community events.
Assets discussed
| Asset | Kind | Stance | Context |
|---|---|---|---|
| AppleAAPL | equity | neutral | Probably Don Valentine. I mean, he wrote the seed check to Apple and NVIDIA. So that is 10 trillion in market cap that is attributed to uh one person. |
| NVIDIANVDA | equity | neutral | Probably Don Valentine. I mean, he wrote the seed check to Apple and NVIDIA. So that is 10 trillion in market cap that is attributed to uh one person. |
“Stance” is how the speakers talked about the asset, not a recommendation.
Claims worth verifying
Metaprop has invested in Attentive Build Ops, Urban, Obi, and Toggle Holdings.
Unidentified speaker (speaker_0) ·0:58
How to verify: Check Metaprop's portfolio page on their website or their LinkedIn profile for investment announcements. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Zach Aarons worked at Millennium Partners, his father's real estate company, around 2010.
Unidentified speaker (speaker_2) ·6:02
How to verify: Check Zach Aarons' LinkedIn profile for employment history at Millennium Partners. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Metaprop was launched in 2015.
Unidentified speaker (speaker_2) ·8:04
How to verify: Check Metaprop's website 'About' section or press releases from 2015. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Metaprop spun up a subsidiary called Metaprop Labs in early 2026.
Unidentified speaker (speaker_2) ·12:53
How to verify: Check Metaprop's press releases or LinkedIn announcements from early 2026 regarding Metaprop Labs. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Handoff and H2 Corporation achieve 97-98% accuracy in takeoffs and estimating, outperforming humans who spot 94% of errors.
Unidentified speaker (speaker_2) ·16:55
How to verify: Review case studies or white papers from Handoff and H2 Corporation regarding their accuracy benchmarks. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Figure pays users to wash dishes and fold laundry to collect training data for world models.
Unidentified speaker (UNKNOWN) ·20:08
How to verify: Check Figure's website or app store description for details on their data collection program. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Fannie Mae and Freddie Mac opened mortgage underwriting to VantageScore after decades of using only FICO.
Unidentified speaker (speaker_0) ·26:51
How to verify: Check press releases from Fannie Mae and Freddie Mac regarding underwriting guidelines and approved scoring models. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Primary sourceBaldwin acquired Obi in January and is going private in a seven billion dollar deal.
Unidentified speaker (speaker_2) ·33:33
How to verify: Check SEC filings (Schedule 13D or proxy statements) for Baldwin Insurance Group and Obi regarding the acquisition and privatization transaction. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Primary sourceSolar costs have seen a tremendous compression.
Unidentified speaker (speaker_2) ·25:34
How to verify: Review Lazard's Levelized Cost of Energy (LCOE) reports or IEA solar energy outlooks for cost trend data. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
China is producing solar panels at a massive scale, with factories the size of cities.
Unidentified speaker (speaker_2) ·26:43
How to verify: Check industry reports from BloombergNEF or IEA on Chinese solar manufacturing capacity and output. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Insurance and accounting are 'triple bullseyes' for AI because they are highly regulated, traditional, and safe.
Unidentified speaker (UNKNOWN) ·33:56
How to verify: Check industry reports on AI adoption in regulated sectors; verify regulatory status of insurance and accounting. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Obi and Baldwin control licenses and distribution, making them less susceptible to disruption by AI agents.
Unidentified speaker (speaker_2) ·34:15
How to verify: Review Obi and Baldwin's business models and regulatory licenses in home improvement. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
C Sequence took a company private because it was trading at a depressed multiple.
Unidentified speaker (UNKNOWN) ·34:50
How to verify: Search SEC filings or press releases for C Sequence (or similar entity) taking a company private around the time of the podcast. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Primary sourceDon Valentine wrote the seed check to Apple and NVIDIA.
Unidentified speaker (speaker_2) ·38:32
How to verify: Check historical VC records or biographies of Don Valentine and Sequoia Capital for seed investments in Apple and NVIDIA. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Primary sourceApple and NVIDIA have a combined market cap of 10 trillion attributed to Don Valentine's seed checks.
Unidentified speaker (speaker_2) ·38:32
How to verify: Check current market cap data for Apple and NVIDIA; verify the attribution claim is a rhetorical estimate rather than a precise financial fact. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Primary sourceMichael Chang was Kei Nishikori's coach.
Unidentified speaker (speaker_2) ·43:27
How to verify: Check tennis coaching records for Kei Nishikori and Michael Chang. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Kei Nishikori's final tournament was in Tokyo.
Unidentified speaker (speaker_2) ·43:27
How to verify: Check tennis tournament schedules and Nishikori's retirement announcements. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
MetaProp is hosting NYC Real Estate Week events during the week of October 19th.
Unidentified speaker (speaker_2) ·44:31
How to verify: Check MetaProp's website or event calendars for October 2026 events. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Primary sourceZach Schweida is an up-and-coming American tennis player who serves at 140 mph.
Unidentified speaker (speaker_3) ·42:49
How to verify: Check ATP tour player profiles for Zach Schweida's nationality and serve speed statistics. For recent events, verify 2026-08-09 through 2026-11-09. Use another period only when the speaker explicitly names it.
Do your own research
- Verify MetaProp's portfolio company valuations and liquidity events against public filings or press releases.
- Check current grid capacity constraints and solar installation forecasts from the EIA or IEA.
- Review Fannie Mae and Freddie Mac underwriting guidelines regarding non-FICO scores and AI-driven cash flow analysis.
- Assess the accuracy claims of specialized AI tools like Handoff or Lens AI against independent benchmarks.
- Investigate the operational status and funding of robotics companies like Figure to verify data collection strategies.
- Analyze the adaptive reuse potential of office buildings by reviewing local zoning laws and data center power requirements.
- Examine the secondary market liquidity trends for venture-backed companies to validate claims about decoupled valuations.
Full transcriptexpand
- 0:00
- Unidentified speaker (speaker_0): This episode is brought to you by Ramp, the spend management platform we use here at TFR. They're offering listeners $150 just to take a demo. We've never had an offer quite like this. Claim your $150 before this offer is gone at our partner link, ramp.com/slash partners slash TFR. Now
- 0:20
- Unidentified speaker (speaker_0): on to the episode.
- 0:23
- Unidentified speaker (speaker_1): Welcome to the podcast about venture capital, where investors and founders alike can learn how VCs make decisions and reach conviction. Your host is Nick Moran, and this is the Full Ratchet.
- 0:38
- Unidentified speaker (speaker_0): Zach Ahrens joins us today from Brooklyn. He's the co-founder and general partner at Metaprop, a seed and gross Growth stage venture firm investing in technology for the built world. Before Metaprop, he worked in banking at a dot com, in VC for ENIAC, and was a project manager at Millennium
- 0:58
- Unidentified speaker (speaker_0): Partners. Metaprop has invested in companies including Attentive Build Ops, Urban, Obi, and Toggle Holdings, amongst others. Zach, welcome to the show.
- 1:07
- Unidentified speaker (speaker_2): Thanks so much for having me, Nick. I'm an avid listener and I'm just absolutely thrilled to be here.
- 1:12
- Unidentified speaker (speaker_0): Awesome. Well, it's great to have you. I know that you are were on your dad's job sites in a hard hat at age eight. Give us the quick story. When did it become clear that you wanted to commit your career to investing at the intersection of tech and real estate?
- 1:25
- Unidentified speaker (speaker_2): It was a long and circuitous journey, not necessarily all deliberate. I grew up my whole life with one foot in the early stage technology world and the the world. the wide web as we called it at the time and and one foot uh on a construction site and that has sort of
- 1:45
- Unidentified speaker (speaker_2): continued uh from a very early age throughout my my adult professional career as well and I I I can't escape it you know just you know it's uh just when I try to get out they suck me right back in. it. So just for background, yeah I grew up in New York City. My father was always in the real estate business. Uh he ran
- 2:05
- Unidentified speaker (speaker_2): what's now NYC EDC on the public side, which uh we're thrilled to actually have that entity as a a limited partner in Metaprop. So it's that was kind of a full circle moment for me. But then he went went into private real estate development and he's been building buildings uh around the country ever ever since then. So
- 2:25
- Unidentified speaker (speaker_2): I had the joy as an eight-year-old kid to get uh up in a hoist with a hard hat, had a bunch of the construction workers saying really inappropriate things to me. And so I I I always sort of fell in love with the physical world, the built environment, the real estate business, the construction business, uh from a
- 2:45
- Unidentified speaker (speaker_2): very, very young age. But I also had other other interests and passions. Keep in mind this was also the time when people were starting to use platforms like AOL, and my friends and I were starting to go online for the first time. I had a very active Geo Cities page, for example, in the 90s. So I
- 3:05
- Unidentified speaker (speaker_2): was always very interested in the connectivity and the technological capabilities of the internet. And growing up in New York City, uh, it's tough to ignore Wall Street. It's tough to ignore the finance industry, right? So, you know, I'd go to my
- 3:25
- Unidentified speaker (speaker_2): friends' houses, and the ones who had the nicest houses, well, what did their parents do? Well, they they were traders, they were on Wall Street, they were in private equity, they were in hedge funds. So my friends and I started devouring all this. literature that was written, not literature, but but but nonfiction that was written in the late 80s and
- 3:45
- Unidentified speaker (speaker_2): early 90s about the industry, you know, books like Liars Poker and Barbarians at the Gate. And so so I always had this sort of trifecta of potential career opportunities I was interested in, whether that be the the the nascent internet
- 4:01
- Unidentified speaker (speaker_2): uh the built environment, and finance. And I'm lucky enough now to do a job that encapsulates all three of those. At Metaprop, I'm a venture capital investor, so I'm uh in the finance industry. Uh I work with early stage uh technology companies
- 4:21
- Unidentified speaker (speaker_2): primarily. So I'm in the technology industry, and also those companies are servicing the physical world and the players uh within it. I've kept uh all of these things sort of consistent over the years, but
- 4:37
- Unidentified speaker (UNKNOWN): just
- 4:37
- Unidentified speaker (speaker_2): to fast forward a little bit on my background, my first internet job was uh an absolute treat. I was a summer intern working for a company called foreleaf.com, 4leaf.com, which got swept up unfortunately in the 2000. Um uh bubble burst. But it was effectively an earlier version of LinkedIn. It was a great idea. It was a professional networking
- 4:57
- Unidentified speaker (speaker_2): site. And that's where I sort of caught the professional startup bug. And I was like, wow, this is a really fun environment to be working in. I can learn a lot really quickly. Um, I went to college at Brown University after college chasing my wall suit dreams. I went uh into investment banking. also learned a lot uh during those
- 5:17
- Unidentified speaker (speaker_2): years, but realized it probably wasn't uh the best fit for me. Um, so I got an offer to uh help my friend start a startup in the tech tourism industry called Travel Goat. And I went into that business, ultimately ran that for years. Uh, we were putting uh walking tours
- 5:37
- Unidentified speaker (speaker_2): uh on the iPod. originally. We were recording audio walking tours, putting them in the iPod. Once the iPhone launched, we had our own app for sort of touring virtually. I built up uh a community of in-person tour guides, became a tour guide myself. I've led over 300 tours of New York City. I'm very, very
- 5:57
- Unidentified speaker (speaker_2): passionate still about it. A couple times a year, come out of retirement and do tours for startup
- 6:01
- Unidentified speaker (UNKNOWN): founders, uh
- 6:02
- Unidentified speaker (speaker_2): limited. partners of ours as well. It's always a great joy for me. Then after that, I started working for my dad at his real estate company, Millennium Partners. He thought early on, this was about 2010, that we could leverage a lot of new content marketing, social media marketing ideas, and bring those into To the real estate
- 6:22
- Unidentified speaker (speaker_2): development, real estate sales, real estate leasing uh process. This was considered.
- 6:27
- Unidentified speaker (speaker_2): very avant-garde 15 years ago. Now social media is ubiquitous within the industry, but there was a very small community of people doing it back then. So that's how I got the bug for the real estate and construction business.
- 6:44
- Unidentified speaker (UNKNOWN): And
- 6:44
- Unidentified speaker (speaker_2): then I sort of had a quarter life crisis of sorts, right? I'm interested in finance. I'm interested in early stage venture and startups. I'm interested in real estate. What do I do with my life? How do I actually translate that into a career that makes sense? So I went to Columbia Business School. My goal was I'm gonna
- 7:04
- Unidentified speaker (speaker_2): figure it out. And I was lucky enough to have a professor there, Stu Ellman, who taught the venture seminar. He's a VC founded R RRE R E uh Ventures, which is one of the oldest. the and and and most sort of august firms we have here in New York. Uh still remains a close friend and mentor of mine. And I was in his class and he suggested that I
- 7:24
- Unidentified speaker (speaker_2): focus on this intersection. He said there was going to be tremendous growth in the industry. And he had looked at my background and he said, You know, you work at a real estate company, you've worked in tech, you've you're doing this venture work for ENIAC. Anyac. You actually have a chance to distinguish yourself. from the rest of the sort of broader general venture community if you focus on
- 7:44
- Unidentified speaker (speaker_2): this and you really thoughtfully put your mind to aggressively growing your reputation here. And that's what I did. Um, I worked in real estate. Um, I was an angel investor in a lot of companies that are look similar to what we're investing in today uh at Metaprop. I was a rabid uh user. pilot tester um of
- 8:04
- Unidentified speaker (speaker_2): all these technologies and that ultimately led to meeting my partners and launching Metaprop in 2015. This was at a time in New York City. We had Deputy Mayor Alicia Glenn. She coined the hyphen tech economy. New York wanted to be the center of fintech. We wanted to be the center of marketing tech. We wanted want to be the center of ed tech.
- 8:24
- Unidentified speaker (speaker_2): And there was really nothing in New York specifically for these sectors for prop tech, real estate tech, the built environment, whatever one calls it. Uh, so we decided to try to fill that void and ultimately start Metaprop in in 2015.
- 8:39
- Unidentified speaker (UNKNOWN): So
- 8:39
- Unidentified speaker (speaker_0): if you were just a real estate investor, what's the next big neighborhood or area in New York that you think is gonna be hot?
- 8:47
- Unidentified speaker (speaker_2): The next big New York neighborhood?
- 8:49
- Unidentified speaker (speaker_0): Yeah. We've
- 8:49
- Unidentified speaker (speaker_2): had quite a run-up. in many neighborhoods hitness over the last decade. I'm gonna go with Motthaven in the Bronx. Been a tremendous amount of development there. I was up there, my son had a soccer game up there about a year ago. And I went around and and and I saw all of the indicators of a burgeoning, uh,
- 9:10
- Unidentified speaker (speaker_2): cool neighborhood. Art galleries, coffee shops, lots. of pedestrian traffic, lots of cool independent boutiques. And development started there. There's there's lots of affordable housing there. There's lots of market rate rental housing there and high-quality buildings. It has unbelievably abundant
- 9:29
- Unidentified speaker (speaker_2): subway access. You can be, if you work in near Grand Central, uh you can be there in 15 minutes, door to door.
- 9:36
- Unidentified speaker (speaker_0): The next neighborhood, you heard it here first from Zach. You know, there's there's a guy running around Chicago that he's been early to like every real estate trend. He's made a lot of money. And I asked him years ago, I'm like, how do you know what's coming next? And he said, I just find where the artists, musicians, and entrepreneurs are hanging out. And that's that's where I
- 9:56
- Unidentified speaker (speaker_0): invest. And that's usually what draws a lot of other people in. But back to Metaprop. So you started Metaprop. The bet was that real estate software moves to the cloud. What's the bet now?
- 10:06
- Unidentified speaker (speaker_2): Yeah, we had we had migration from on-prem to cloud. We had an explosion of data. Now it's taking all that data and taking all the cloud solutions that house all that data and building AI applications on top of it all that give the customer unprecedented power
- 10:26
- Unidentified speaker (speaker_2): efficiency. and control. I think I've heard it on many podcasts, including this one, the difference between static SaaS and AI is that AI gives the end customer, in our case, that could be a property manager, that could be a construction scheduler, that could be an insurance underwriter. It
- 10:46
- Unidentified speaker (speaker_2): gives them control of their own product and their own workflow. And then this is truly unprecedented. So we're investing primarily at the application layer as it relates to these technologies. We're seeing rapid customer adoption. The company, the the companies we deal with, our limited
- 11:06
- Unidentified speaker (speaker_2): partner base, they are rabid for new technology that can increase efficiency and throughput. And I know there's an article every day about how. everybody's adopting, but no one's seeing efficiency gains. I don't, I'd love to talk to those people because
- 11:21
- Unidentified speaker (speaker_0): everybody
- 11:22
- Unidentified speaker (speaker_2): I talk to, while they're certainly frustrated that they're not moving faster, maybe their competitors doing something they deem is cooler. Every single person I talk to in the industries we cover, uh, insurance, finance, real estate, construction, and they are seeing a tremendous amount of efficiency gain from deployments. deploying some of these new generative AI tools.
- 11:41
- Unidentified speaker (speaker_0): Let's talk about some of the incumbent operators, right? There's a lot of people that think large companies are going to be writing their own software in a few years, you know, with the help of models. Where do you stand on this? You know, the startups, the techs, the the tech companies, the uh up and comers versus the
- 12:01
- Unidentified speaker (speaker_0): established businesses. that have data sets that in theory could do this themselves.
- 12:07
- Unidentified speaker (UNKNOWN): Yeah,
- 12:07
- Unidentified speaker (speaker_2): I think there will be a few very large players and then a few fast moving startup players that are actually vibe coding their own full-blown software solutions, vibe coding their own serinum, vibe coding their own ERP equivalent. It's very easy to vibe code something with today's technology.
- 12:27
- Unidentified speaker (speaker_2): It's very challenging to vibe maintain something.
- 12:31
- Unidentified speaker (speaker_0): That's right. Or optimize.
- 12:33
- Unidentified speaker (speaker_2): We saw it just uh just the other day. We've spun up a subsidiary called Metaprop Labs that's an AI enablement and deployment company for the industry. We see all the time a new package gets shipped, an API changes, and the whole thing breaks down. And the these folks at the incumbent firms
- 12:53
- Unidentified speaker (speaker_2): and the mid-market firms, they don't have the internal teams that are knowledgeable enough and capable enough and have enough time on their hands, frankly, to maintain maintenance. So, a couple trends we saw early 2026 that we've already seen mean reversion from. One of them is that everybody's gonna vibe code their own thing. No, sorry,
- 13:13
- Unidentified speaker (speaker_2): these firms that have been around for 20 years are agentifying or they're dying. And once they agenfy, the customer's gonna rather use the agentic version of that platform that has maybe an MCP server, has all the APIs linked up, then they are to vibe code their own. The other thing we saw, right, was unbridled token spend. So
- 13:33
- Unidentified speaker (speaker_2): CFOs finally got the token bill. I think it hit sometime this summer, and they said, Whoa. whoa, whoa. This is supposed to be cheaper than humans. Yeah. Otherwise, let me just hire a bunch of humans. Maybe they'll do it slower.
- 13:48
- Unidentified speaker (UNKNOWN): The
- 13:49
- Unidentified speaker (speaker_2): cost ran out of control. So we're seeing now a radical, and I think this is great for the industry, a radical um sophistication shift where everybody is now very conscious of compute. Um, they're asking their employees to tell them why they need to go above and
- 14:09
- Unidentified speaker (speaker_2): beyond and use additional compute. More sophisticated companies are building their own harnesses so they can leverage different models for different tasks. You don't need
- 14:20
- Unidentified speaker (speaker_2): Mythos to
- 14:22
- Unidentified speaker (speaker_2): get a recipe for Arctic Char. Right?
- 14:27
- Unidentified speaker (speaker_2): Just don't. And so, and we're also seeing new models getting built every day. For example, we're working on incubating commercial mortgage servicing company. And I was talking with the entrepreneur, and he said, I'm thinking of building the entire thing using Jev. Jev didn't even exist three months ago. But for his purposes with in commercial
- 14:47
- Unidentified speaker (speaker_2): loan servicing, he doesn't need the latest and greatest experience. expensive compute model coming out of one of the better known foundations.
- 14:55
- Unidentified speaker (speaker_0): When you look at these companies that these AI companies that are selling into real estate or construction, how do you get comfortable that it's still going to be around after the next model release, you know, after these things continue to to innovate and evolve?
- 15:12
- Unidentified speaker (speaker_2): Well you can never be totally sure. We don't we don't have necessarily we have friends at the foundation model companies, but we don't really know where their priorities are, right? So you're ultimately doing a bit of guesswork. However, there are lots of situations, and I'll focus on one area in particular where we've invested tremendous
- 15:32
- Unidentified speaker (speaker_2): amounts of capital in multiple companies. There are certain areas that require a combination of determination. deterministic and probabilistic. And the deterministic rails rely on very advanced, typically computer vision technology. And in our
- 15:51
- Unidentified speaker (speaker_2): industries, there's a tremendous amount of documentation. There are millions of pages of spec sheets, drawing sets coming from multiple different constituencies. that never match up. If you throw those million pages even into the most advanced foundation model, the one that can solve
- 16:11
- Unidentified speaker (speaker_2): Navier Stokes, the one that can um hack into government, you know, uh uh infrastructure across the world, they will effectively puke all over these documents, or they will deliver something. that's roughly 65 to 70%
- 16:30
- Unidentified speaker (speaker_2): accurate compared to a benchmark against a human team.
- 16:35
- Unidentified speaker (UNKNOWN): So
- 16:35
- Unidentified speaker (speaker_2): we've invested in companies like Handoff on the takeoff and estimating front, H2 Corporation in in Japan doing takeoffs and estimating, Light Table, which is doing uh peer review, which is a company we co-incubated with primary ventures. These folks. are at 97 to 98% accuracy.
- 16:55
- Unidentified speaker (speaker_2): And they're almost completely getting rid of human in the loop. When humans look at these, if if you give a human a million pages of specs and drawings, and you say, look at all the errors, they will usually spot 94% of them.
- 17:10
- Unidentified speaker (UNKNOWN): And
- 17:11
- Unidentified speaker (speaker_2): it'll take them, by the way, 20 to 30 times longer.
- 17:15
- Unidentified speaker (UNKNOWN): than
- 17:16
- Unidentified speaker (speaker_2): it takes the AI engine. So that's one category where we can feel very confident that we have a significant leg up over the foundation model companies. And there's also tremendous demand from the industry for these solutions right now.
- 17:32
- Unidentified speaker (speaker_0): Interesting. Talk talk about on-site on site applications. So I'm thinking like real-world real world applications. application, not just software, but robots, role models, you know, those are getting a lot of attention. Uh as we think about job sites or, you know, inside a building, um, where do you expect
- 17:53
- Unidentified speaker (speaker_0): robotics and physical applications to show up first?
- 17:57
- Unidentified speaker (speaker_2): They're already showing up
- 17:59
- Unidentified speaker (UNKNOWN): on
- 17:59
- Unidentified speaker (speaker_2): many different job sites, most notably Solar Farm. When you have a solar farm, you have a basically flat plane and you have a very repeatable process for mounting solar arrays.
- 18:13
- Unidentified speaker (UNKNOWN): So
- 18:13
- Unidentified speaker (speaker_2): we've looked a lot. Um
- 18:17
- Unidentified speaker (speaker_2): we've actually bid on a couple rounds of companies that are looking to deploy autonomous or semi-autonomous robots mounting solar arrays.
- 18:27
- Unidentified speaker (UNKNOWN): But
- 18:27
- Unidentified speaker (speaker_2): we're seeing it everywhere. Robots are now laying bricks, drywall, uh, their robotic crane operators, robotic excavators. I think one thing people get wrong
- 18:41
- Unidentified speaker (UNKNOWN): in
- 18:41
- Unidentified speaker (speaker_2): their assumption about this is that it's going from like zero to full autonomy tomorrow. That's not how it works.
- 18:48
- Unidentified speaker (UNKNOWN): It
- 18:48
- Unidentified speaker (speaker_2): These are telling teleoperations.
- 18:50
- Unidentified speaker (UNKNOWN): fake
- 18:50
- Unidentified speaker (speaker_0): figures.
- 18:53
- Unidentified speaker (speaker_0): Well,
- 18:53
- Unidentified speaker (speaker_2): figure's different. I'm gonna I can I can get to that, but but figures figures a little different. I'm talking just about construction robotics. The vast majority of the companies right now
- 19:04
- Unidentified speaker (speaker_2): are teleoperated.
- 19:06
- Unidentified speaker (UNKNOWN): If you
- 19:06
- Unidentified speaker (speaker_2): think about Waymo, they have a call center in the Philippines. If things get really bad with the autonomous vehicle, a human can intervene. We're seeing the same thing. deployed onto the job sites. Companies like Figure
- 19:19
- Unidentified speaker (speaker_2): represent a very interesting opportunity to put robots in everybody's home.
- 19:25
- Unidentified speaker (speaker_2): Now that's going to be
- 19:27
- Unidentified speaker (UNKNOWN): a
- 19:28
- Unidentified speaker (speaker_2): wave that follows, right? We had robots in manufacturing and warehousing logistics. That's almost ubiquitous at this point.
- 19:38
- Unidentified speaker (UNKNOWN): Now
- 19:39
- Unidentified speaker (speaker_2): we're gonna have robots on the job site. Then we're gonna have robots in the home.
- 19:44
- Unidentified speaker (speaker_2): Concierge of the hotel is gonna be a robot.
- 19:48
- Unidentified speaker (UNKNOWN): One
- 19:48
- Unidentified speaker (speaker_2): of my favorite things about figures specifically is that and ties into world models, right? They're building their own world model, and they don't have enough data to train the robot to do the household task because almost every household is different and the configurations are unique.
- 20:08
- Unidentified speaker (UNKNOWN): So
- 20:08
- Unidentified speaker (speaker_2): they're actually, they've launched an app and a device where they'll come to your house and video you washing dishes and pay you.
- 20:15
- Unidentified speaker (UNKNOWN): I
- 20:16
- Unidentified speaker (speaker_2): don't know if you've seen this.
- 20:17
- Unidentified speaker (speaker_0): Uh no. But
- 20:18
- Unidentified speaker (speaker_2): yeah, figure, I'm still on the wait list. Figure will pay you to wash dishes, fold laundry, um, sweep the floor because they want to get a sense of how to train their robots. And you're seeing a lot of interesting the training's the most interesting component to me, right? We saw the AMD World Labs acquisition. Companies like Generalist are super
- 20:38
- Unidentified speaker (speaker_2): interesting where you have robots sort of training other robots through video interface. Um, this is all brand new territory. Uh, it's an area we're excited to figure out how Metaprop invests, right? We're not necessarily going to put $200 million out of a hundred million dollar fund into a world lab company, but we want exposure.
- 20:58
- Unidentified speaker (speaker_0): Let me double click on this. So I'm, I have to admit, I'm a little bearish on figure near term. I think long term, you know, it can be fantastic. But I think the applications are really difficult. And I think they're varied and I think there's a number of them. But if we look at figure and it's trying to address Washington washing District. dish dishes or folding laundry, I think
- 21:17
- Unidentified speaker (speaker_0): that that problem is actually in a more controlled environment and a little bit more repeatable than like a job site where I've been to some job sites, not as many as you, but it can be chaos and there's surprises, right? You dig up the ground, you find something you didn't expect, or you know,
- 21:37
- Unidentified speaker (speaker_0): there's weather. or the wrong sub showed up at the wrong time, right? And like all of a sudden that crashed schedule that you had and that critical path is completely blown up. Yeah. Because you can't do these things in serial. So, like, help us understand, you know, how
- 21:57
- Unidentified speaker (speaker_0): robotics relate to a site that's, you know, very unpredictable. In
- 22:02
- Unidentified speaker (speaker_2): the construction. context, the robots do one specific task over and over again, supervised by a human in a controlled environment. Right? The leverage is that one human can supervise five robots,
- 22:14
- Unidentified speaker (UNKNOWN): and
- 22:14
- Unidentified speaker (speaker_2): that job will be done theoretically more accurately and much faster. So there's an actual labor savings. The problem with construction robots, there there've have been construction robots since we started Metaprop. The problem in the past. is that ratio has been off. No one's been able to deploy enough robots to really make
- 22:34
- Unidentified speaker (speaker_2): it worth anybody's while. But that is what's now changing the the capabilities of the robots are getting better, they're getting more reliable, they're getting cheaper and easier to produce. But these you're not gonna see a robotic foreman.
- 22:50
- Unidentified speaker (speaker_2): on a job site for a very long time. The issues that pop up, yes, those those those for a while are gonna be handled by humans. Absolutely. So,
- 23:00
- Unidentified speaker (speaker_0): Zach, you wrote the summer that AI's real constraint is no longer the chip. What is the new constraint?
- 23:06
- Unidentified speaker (speaker_2): The biggest constraint is power. We really don't have enough grid power for all these data centers. There's a big promise. with geothermal. There's a big promise with solar. There's a big promise with nuclear.
- 23:20
- Unidentified speaker (speaker_0): Co-located natural gas.
- 23:22
- Unidentified speaker (speaker_2): Natural gas as well.
- 23:23
- Unidentified speaker (UNKNOWN): Like
- 23:23
- Unidentified speaker (speaker_0): Bloom is I think Bloom is actually
- 23:26
- Unidentified speaker (speaker_2): yeah. So if you look at
- 23:27
- Unidentified speaker (speaker_0): the sponsor of the NYX now, right?
- 23:28
- Unidentified speaker (speaker_2): Bloom has had its stock price run up massively. We don't know if all those are gonna functionally get delivered to the job site in time.
- 23:37
- Unidentified speaker (UNKNOWN): Right?
- 23:37
- Unidentified speaker (speaker_2): The build build-outs outs scheduled to peak in 2028. How ready is a SMR company going to be?
- 23:45
- Unidentified speaker (UNKNOWN): Right?
- 23:46
- Unidentified speaker (speaker_2): Geothermal lives or dies by this specific project. You have a geothermal site. If that data center is delayed, that geothermal company risks um a cash crunch. Right. So you have all of these factors and then you also have increasing political
- 24:06
- Unidentified speaker (speaker_2): uh pushback against
- 24:09
- Unidentified speaker (UNKNOWN): some
- 24:09
- Unidentified speaker (speaker_2): of these large AI infrastructure build-outs. And you have the promise of orbital and oceanic data centers, but with orbital, we're still a few years out from that, from the cost compressing enough to make it really worthwhile. And for the big training,
- 24:25
- Unidentified speaker (UNKNOWN): we
- 24:26
- Unidentified speaker (speaker_2): really need to. terrestrial data centers. We're not going to put that infrastructure necessarily up in space. And so I view that by far as the biggest uh constraint.
- 24:37
- Unidentified speaker (UNKNOWN): Um, I
- 24:38
- Unidentified speaker (speaker_2): know everyone says there's this, you know, there is a huge adoption gap. I think that's gonna be the fastest to close. I think the
- 24:49
- Unidentified speaker (UNKNOWN): one
- 24:49
- Unidentified speaker (speaker_2): of the things we saw. in 26 was people getting a lot smarter about deployment and enablement. And you saw these consultancies, private equity firms popping up saying, We're gonna
- 25:01
- Unidentified speaker (UNKNOWN): put
- 25:01
- Unidentified speaker (speaker_2): forward deployed engineers on your site. You're gonna get it. You're gonna get all the capability, right? We're gonna make sure you get it before we before we leave, or you were they're gonna stay with you forever. The biggest bottleneck is definitely energy. and will remain so.
- 25:16
- Unidentified speaker (speaker_0): I know that there's lots of forms of energy, and a lot of people are saying we need to make it up in the aggregate. Is is there one form that you feel like has enough speed, modularity, etc., that it's it's gonna become sort of a a dominant player in the data center in AI build-out? build-up.
- 25:34
- Unidentified speaker (speaker_2): We've seen a tremendous compression of cost for. solar. Solar is abundant, right? Yeah.
- 25:42
- Unidentified speaker (UNKNOWN): We'll
- 25:42
- Unidentified speaker (speaker_2): talk about nuclear, but solar is nuclear.
- 25:47
- Unidentified speaker (UNKNOWN): Right?
- 25:47
- Unidentified speaker (speaker_2): The sun is a gigantic nuclear weapon that is is is is giving energy to this to this planet, right? And so
- 25:56
- Unidentified speaker (UNKNOWN): I
- 25:56
- Unidentified speaker (speaker_2): think solar doesn't have political opposition
- 26:01
- Unidentified speaker (UNKNOWN): in
- 26:01
- Unidentified speaker (speaker_2): the same way.
- 26:03
- Unidentified speaker (speaker_2): that natural gas, you know, turbines have, or electricity that's from the grid that's coming from a coal plant, right? Um doesn't necessarily have the same risks as fission fusion on the nuclear side, um, and isn't bound to any specific
- 26:23
- Unidentified speaker (speaker_2): underground location like geothermal. So I think fundamentally, I think solar is going to power the bulk of the AI build out going forward. I mean, if you look at if you look at especially China, like all of China's just ripping so many solar panels off these gigantic factories. They have factories the size of cities,
- 26:43
- Unidentified speaker (speaker_2): just cranking these things
- 26:45
- Unidentified speaker (UNKNOWN): out.
- 26:45
- Unidentified speaker (speaker_2): And all of their AI infrastructure is ultimately going to be powered by solar. Interesting.
- 26:51
- Unidentified speaker (speaker_0): So switching gears here, Fanny and Freddie just opened mortgage underwriting to vantage score after decades of using FICO only. Um, what does that change for the companies you look at in housing?
- 27:04
- Unidentified speaker (speaker_2): We have a number of investments in companies that have developed their own AI underwriting tools that we think outperform uh the FICO score. So SplitPay, which is uh one of our fastest growing companies we've ever backed, that company's developed a platform called Lens AI, which can underwrite much faster,
- 27:24
- Unidentified speaker (speaker_2): very, very accurate, much less, much easier to catch fraud. And we think that Fannie Freddie and a number of of other government agencies are going to open up beyond the now two approved parties, that within a few years they'll probably be
- 27:44
- Unidentified speaker (speaker_2): 15 to 20 that an underwriter um and that has applications we think well beyond just uh just mortgages, also in in in renting apartment, Department, also in um the insurance business, also in the auto. business.
- 28:04
- Unidentified speaker (speaker_2): Um, so we're very excited to see that the FICO sort of monopoly is is is potentially cracking a little bit because we do believe there are new agentic ways to better underwrite credit quality than what's currently off.
- 28:21
- Unidentified speaker (speaker_0): What signals are they looking at? What are the underlying data sets that they're pulling that FICO's missing?
- 28:25
- Unidentified speaker (speaker_2): They're mainly looking at your ability. to pay cash inflows and cash outflows. They're less looking when when you look at a FICO score, um, somebody can have a relatively good FICO score because they've had really good credit repayment in the past, but be on the verge of personal
- 28:44
- Unidentified speaker (speaker_2): insolvency, um, just from a cash perspective. So it really boils down to that. Are you
- 28:51
- Unidentified speaker (speaker_0): still managing expenses with outdated apps and spreadsheets? If so, it's time for Ramp, the leading US spend management platform covering corporate cards, travel, AP, and more. All built to save time and extend runway. We use Ramp at TFR and it's been a game changer. As our preferred partner, they're offering
- 29:10
- Unidentified speaker (speaker_0): listeners $150 just to take a demo. No need to sign up. Head to ramp.com/slash partnerslash TFR to claim your $150 before the offer expires. Now back to the interview.
- 29:24
- Unidentified speaker (speaker_3): Where do you think um technology
- 29:27
- Unidentified speaker (speaker_0): can kind of move the needle on affordability when it comes
- 29:30
- Unidentified speaker (UNKNOWN): to uh
- 29:31
- Unidentified speaker (speaker_0): owning a home and and uh
- 29:35
- Unidentified speaker (UNKNOWN): you
- 29:35
- Unidentified speaker (speaker_0): know living situations?
- 29:37
- Unidentified speaker (UNKNOWN): Yeah,
- 29:37
- Unidentified speaker (speaker_2): I think fundamentally the big biggest impact we can have is through material sciences. So if we can really, because if you look at the cost of lumber, cost of steel, inputs that go into that,
- 29:49
- Unidentified speaker (UNKNOWN): we
- 29:50
- Unidentified speaker (speaker_2): just had runaway inflation uh for a very, very long time. And that is ultimately squeezing affordability tremendously. We still have a massive amount of housing that needs to be. built in this country, you know, roughly million units. Uh
- 30:09
- Unidentified speaker (UNKNOWN): but
- 30:09
- Unidentified speaker (speaker_2): if we don't have better, cheaper, and faster building materials
- 30:14
- Unidentified speaker (UNKNOWN): than
- 30:14
- Unidentified speaker (speaker_2): we currently have, where where there's really been very limited uh innovation.
- 30:19
- Unidentified speaker (UNKNOWN): Um
- 30:20
- Unidentified speaker (speaker_2): graphene, I don't know if you ever looked into graphene. That was supposed to be our savior. Yeah. You don't didn't even know about it. It it totally petered out, right? And but one of the benefits now of having an AI model that can solve Navier Stokes is you can have an AI model come up with material
- 30:40
- Unidentified speaker (speaker_2): composites that we've never seen before. So we have a material science and 3D printing company in Chattanooga, Tennessee called Branch Technologies. They're working with,
- 30:49
- Unidentified speaker (UNKNOWN): they're
- 30:49
- Unidentified speaker (speaker_2): trying to manufacture ultimately. what's going to go up when we have housing in space, right? What is that going to look like working with complex polymers? They are playing with the foundation model tools to see, okay, how do we really push the envelope and develop materials that that were sort of beyond the the capabilities of the human
- 31:09
- Unidentified speaker (speaker_2): mind to even ideate on?
- 31:11
- Unidentified speaker (speaker_0): A question I keep getting from folks is um
- 31:15
- Unidentified speaker (speaker_0): with the pace of innovation, with how rapid companies can move, like you're you're looking or you've invested in a company that's looking at you know material building materials for for space domiciles. Like that's wild to me. But the the question I keep getting is because of the pace of innovation and how much that can speed
- 31:35
- Unidentified speaker (speaker_0): up with AI and
- 31:38
- Unidentified speaker (UNKNOWN): the
- 31:38
- Unidentified speaker (speaker_0): lower demand on talent and cost and some other things that were embedded maybe in previous supercycles. Should we expect a much faster ramp of these companies? I mean, we're already seeing growth rates that are unprecedented, but will that compress venture fund
- 31:58
- Unidentified speaker (speaker_0): cycles? Will that speed up DPI? I mean, I don't want to make any problems. promises to my LPs, but it's a very reasonable question.
- 32:07
- Unidentified speaker (speaker_2): I think it should compress venture cycles. However,
- 32:13
- Unidentified speaker (UNKNOWN): the
- 32:13
- Unidentified speaker (speaker_2): pushback I would give you on that is while those dynamics are happening, right? We have companies growing from zero to 100 million in 12 to 18 months, and we've seen it even within our own portfolio.
- 32:25
- Unidentified speaker (UNKNOWN): Those
- 32:26
- Unidentified speaker (speaker_2): companies are also still staying private for longer. There's still scarcity in the market. So those companies' valuations are getting massively pushed up to multiples that are now decoupled from similar companies in the public equity markets.
- 32:46
- Unidentified speaker (speaker_2): Um, So while the
- 32:49
- Unidentified speaker (UNKNOWN): the
- 32:49
- Unidentified speaker (speaker_2): growth and the speed and the velocity should help the venture ecosystem, I think all of those other things are are pressing against that growth. And it might sort of even out
- 33:03
- Unidentified speaker (UNKNOWN): interest in the
- 33:03
- Unidentified speaker (speaker_2): next few years. I think the benefit is companies, yes, are staying private longer It's and longer, maybe seeking MA less and less, but there's a much more active secondary market. So if you're running a venture fund, there are many, many more opportunities for liquidity on the secondary market than there were 10 years ago when we started the firm.
- 33:24
- Unidentified speaker (speaker_0): I wanted your quick take on insurance too. I know that Baldwin bought Obi in January, and now Baldwin itself is going public in a seven billion dollar deal. Going
- 33:33
- Unidentified speaker (speaker_2): private.
- 33:33
- Unidentified speaker (UNKNOWN): Or
- 33:34
- Unidentified speaker (speaker_0): sorry, going private.
- 33:36
- Unidentified speaker (UNKNOWN): Um,
- 33:36
- Unidentified speaker (speaker_0): I'm reading too fast here. But you know, what did you learn about that series of events about kind of how insurance businesses are being valued?
- 33:45
- Unidentified speaker (speaker_2): If you look at very, very large markets that have been traditionally analog, where there's a very obvious application
- 33:56
- Unidentified speaker (UNKNOWN): for
- 33:56
- Unidentified speaker (speaker_2): a gentic solutions, insurance is like a triple bullseye. Accounting, right? If you look at from like Thrive Holdings, going buying up accounting firms, and I like accounting, also a triple bullseye. Also highly regulated, and per your question
- 34:15
- Unidentified speaker (speaker_2): earlier, relatively safe. If you control the distribution, like an Obi or a Baldwin, control the licenses,
- 34:25
- Unidentified speaker (UNKNOWN): you
- 34:25
- Unidentified speaker (speaker_2): are not necessarily as easy. easily disrupted by an upstart as you might if you were building agents for those businesses.
- 34:34
- Unidentified speaker (UNKNOWN): So
- 34:34
- Unidentified speaker (speaker_2): I think that's behind the motivation. I think Baldwin saw what Obi was doing on the AI front, wanted exposure to that, wanted to disseminate some of that AI sprinkle magic into their other subsidiaries. And then I think
- 34:50
- Unidentified speaker (UNKNOWN): what's what C
- 34:50
- Unidentified speaker (speaker_2): Sequence saw taking it private again was this company's trading at a depressed multiple.
- 34:57
- Unidentified speaker (UNKNOWN): The
- 34:58
- Unidentified speaker (speaker_2): market doesn't understand this. We can further identify this business and we can grow our EBITDA margin by 20% over the next five years.
- 35:04
- Unidentified speaker (speaker_0): Wow. Big impacts coming for insurance.
- 35:07
- Unidentified speaker (UNKNOWN): Yeah,
- 35:08
- Unidentified speaker (speaker_2): we're still very, very active in insure tech. Always been a big focus for us, but we see tremendous opportunity there. You
- 35:14
- Unidentified speaker (speaker_0): know, I know you're writing seed checks, you're also writing series C checks. Like, help me understand the barbell here, you know, why you're entering so early and and later.
- 35:24
- Unidentified speaker (speaker_2): We enter early. We we we think our team is really good at working with early stage entrepreneurs to find product market fit. We think seed and pre-seed are less competitive than series A. By the time the companies have grown to that stage, and we're not
- 35:39
- Unidentified speaker (speaker_3): series B and C's like we would a C. In many cases, we have six, seven years experience with these companies.
- 35:44
- Unidentified speaker (speaker_2): And so we can become comfortable writing the growth check. Whereas Series A to lead it, very competitive, still a lot of risk. So I think we've focused on this barbell strategy
- 35:57
- Unidentified speaker (UNKNOWN): Um,
- 35:57
- Unidentified speaker (speaker_2): to drive significant alpha at the C sea level and then reduce risk. once we have our long-term relationship, can see metrics going for years and have the access because we've now been in the industry so long and know somebody these entrepreneurs.
- 36:13
- Unidentified speaker (speaker_0): How is your thought process and your frame on underwriting it series C? How has that impacted the way you think about seed deals?
- 36:22
- Unidentified speaker (speaker_2): If there's no downstream capital available to you and you're a seed investor, doesn't matter how good the business is performing. It's really hard to do that though, because it's like, will there be downstream capital in three years or four years or whenever it is? But that's a f that's the the the fundamental question I would say we ask at seed every time now, since
- 36:42
- Unidentified speaker (speaker_2): having a uh a growth vehicle that you know we we didn't ask at investment committee,
- 36:47
- Unidentified speaker (UNKNOWN): you know,
- 36:47
- Unidentified speaker (speaker_2): 10 years ago.
- 36:48
- Unidentified speaker (UNKNOWN): Zach,
- 36:49
- Unidentified speaker (speaker_0): give me your boldest non non-consensus prediction about real estate five years from now.
- 36:55
- Unidentified speaker (speaker_2): My boldest non-consensus prediction about real estate is
- 37:03
- Unidentified speaker (speaker_2): I think
- 37:04
- Unidentified speaker (speaker_2): that we are going to see a huge uh rebound in office. We've already started
- 37:13
- Unidentified speaker (UNKNOWN): to
- 37:14
- Unidentified speaker (speaker_2): see it. It's only really impacted the top-tier office in the top-tier cities. I think as we leverage AI, we are going to find adaptive reuse capabilities for even laggard office buildings that we couldn't even think of before. For. I
- 37:33
- Unidentified speaker (speaker_2): think we're going to be able to transform office buildings much quicker into assets like data centers, into things like vertical warehouse and logistics facilities, into self-driving car garages. And these buildings will ultimately be worth much more than people think they're
- 37:53
- Unidentified speaker (speaker_2): worth today. We've already seen an explosion with some successes, some failures of office to residential conversion.
- 38:03
- Unidentified speaker (UNKNOWN): I
- 38:03
- Unidentified speaker (speaker_2): think we'll continue to see that, but we're going to see a lot of activity, office to better office, office to logistics, office to hospitality, office to data center. We just need the political will
- 38:18
- Unidentified speaker (UNKNOWN): to
- 38:19
- Unidentified speaker (speaker_2): get behind that. And we need the technology to advance more so that these things can be economically viable. But I legitimately think we're five years away from that.
- 38:29
- Unidentified speaker (speaker_0): Zach, who are your favorite old school VCs and why?
- 38:32
- Unidentified speaker (speaker_2): Probably Don Valentine. I mean, he wrote the seed check to Apple and NVIDIA. So that is 10 trillion in market cap that is attributed to uh one person. So I would have to go uh
- 38:49
- Unidentified speaker (UNKNOWN): I
- 38:49
- Unidentified speaker (speaker_2): would have to go with him. I just I just think he must have had
- 38:54
- Unidentified speaker (UNKNOWN): a
- 38:54
- Unidentified speaker (speaker_2): tremendous uh vision and he also had great access to great entrepreneurs and he had operating experience too.
- 39:00
- Unidentified speaker (speaker_0): What's the worst advice you've ever gotten? Did you take
- 39:04
- Unidentified speaker (UNKNOWN): it?
- 39:05
- Unidentified speaker (speaker_2): I did.
- 39:07
- Unidentified speaker (speaker_0): What was it?
- 39:10
- Unidentified speaker (speaker_2): So when I was like interviewing
- 39:13
- Unidentified speaker (UNKNOWN): for
- 39:13
- Unidentified speaker (speaker_2): jobs when I was a senior in college, I got advice from a mentor that was like, just tell them you want to get your foot in the door. You don't care what you're gonna do. You're gonna scrub the floors with a toothbrush. And they just value hard work. And that turned out to be the completely wrong advice.
- 39:33
- Unidentified speaker (speaker_2): So what I tell my students when they ask for career advice and recruiting advice is you need to have a perspective and a point of view that makes the interviewer interested in you and want to bring you onto their team. Doesn't have to be directly even related to what they do. You need to show you understand what they do, but they also want you to have a perspective because
- 39:53
- Unidentified speaker (speaker_2): why else are they hiring you? And instead of just saying you'll scrub the floors with a toothbrush, which makes you sound pathetic and desperate,
- 40:01
- Unidentified speaker (UNKNOWN): you
- 40:02
- Unidentified speaker (speaker_2): need to commit communicate an intense and hard work ethic, maybe through telling a story
- 40:07
- Unidentified speaker (UNKNOWN): of
- 40:08
- Unidentified speaker (speaker_2): something you worked on rather than sort of saying you'll do anything just to get a job.
- 40:14
- Unidentified speaker (speaker_0): Zach, if we can feature anyone here on the show, who do you think we should interview and what topic would you like to hear them speak about?
- 40:20
- Unidentified speaker (speaker_2): I would love to hear from Molly Fowler. Have you have you had Molly on the podcast? yet?
- 40:27
- Unidentified speaker (speaker_0): I have not. No. She
- 40:28
- Unidentified speaker (speaker_2): runs Dormroom Fund and she's built this whole infrastructure and very successful investment career out of funding, not just funding, but creating community of college students who are interested in starting companies, being adventure. Um, and she's very, very focused on the community and less on the sort of
- 40:48
- Unidentified speaker (speaker_2): specific ideas necessarily coming out. of it. But she's been extraordinarily successful and has built a model that no one else has been able to replicate, has even maybe tried to people have tried in pockets to replicate it. Um, but it's now ubiquitous. I mean, they're at all these colleges, they're cranking out
- 41:08
- Unidentified speaker (speaker_2): founders at a record clip.
- 41:11
- Unidentified speaker (speaker_0): Love that.
- 41:12
- Unidentified speaker (UNKNOWN): Uh,
- 41:13
- Unidentified speaker (speaker_0): Zach, what book, article, or video? would you recommend to listeners?
- 41:16
- Unidentified speaker (speaker_2): I mentioned them earlier, but I love these old Wall Street books. I love Barbarians at the Gate, Den of Thieves, Um, Liars Pokes.
- 41:26
- Unidentified speaker (speaker_2): So all these 80s, nine, eighties Wall Street. Uh that was really the the the peak of exuberance and craziness on Wall Street will never be. replicated again. Um, it was a a window into a different time.
- 41:45
- Unidentified speaker (UNKNOWN): Um,
- 41:46
- Unidentified speaker (speaker_2): and fundamentally, while there's a lot of bad behavior in those books, which is entertaining, there's also a lot of interesting financial information that could be uh gleaned from reading those.
- 41:59
- Unidentified speaker (speaker_0): Some of the best. Uh, Zach, do you have any habits or behaviors that are a secret? weapon?
- 42:03
- Unidentified speaker (speaker_2): Wow.
- 42:05
- Unidentified speaker (speaker_2): I would have to say I play a lot of tennis. And that is a secret weapon because I can always invite maybe a founder or an LP that plays. Um, and I always have something. If someone else is interested in tennis, I can I always have something to fall back on that's not just uh that's not just business.
- 42:25
- Unidentified speaker (UNKNOWN): Who's
- 42:26
- Unidentified speaker (speaker_0): your favorite? tennis player right now?
- 42:29
- Unidentified speaker (speaker_2): My favorite player right now.
- 42:31
- Unidentified speaker (speaker_0): Yeah. Yeah.
- 42:32
- Unidentified speaker (UNKNOWN): I
- 42:33
- Unidentified speaker (speaker_2): really like this guy, Zach Schweida. He's kind of young, up and coming American on the tour. Um, he's my height. We are not tall. I know you can't tell sort of how tall I am. Um, but he can still serve it 140 miles an hour.
- 42:49
- Unidentified speaker (speaker_3): Whoa.
- 42:50
- Unidentified speaker (UNKNOWN): And
- 42:50
- Unidentified speaker (speaker_2): uh he's got a really, really interesting game. Look out for him uh on the tour in in in the years to come. But um historically, my favorite player is Michael Chan.
- 43:02
- Unidentified speaker (speaker_0): Michael Chang.
- 43:03
- Unidentified speaker (speaker_2): And I got we do a lot of business in Japan.
- 43:07
- Unidentified speaker (UNKNOWN): We
- 43:08
- Unidentified speaker (speaker_2): actually have an IPO uh coming up there at the end of this uh this month with a lot of limited partners over there and when I just went over there, I had the privilege to go to the tennis tournament in Tokyo. I saw Alcaraz playing Yuri Lehechka. I got to see Kay Nishikori's career celebration. That was
- 43:27
- Unidentified speaker (speaker_2): his last tournament ever. He's the greatest Japanese tennis player of all time. And Michael Chang, who used to be his coach, stood up and gave a speech. And that was so
- 43:37
- Unidentified speaker (speaker_0): cool.
- 43:38
- Unidentified speaker (speaker_2): Unbelievable. experience for me because I wasn't even didn't even
- 43:42
- Unidentified speaker (UNKNOWN): put
- 43:42
- Unidentified speaker (speaker_2): it together that he might uh be there in attendance.
- 43:46
- Unidentified speaker (speaker_0): So very quickly, why is the company going public in Japan instead of here?
- 43:50
- Unidentified speaker (UNKNOWN): It's
- 43:51
- Unidentified speaker (speaker_2): a Japanese company.
- 43:53
- Unidentified speaker (speaker_0): Uh okay.
- 43:55
- Unidentified speaker (UNKNOWN): All
- 43:56
- Unidentified speaker (speaker_0): right, there you go. Um and finally here, Zach, what is the best way for listeners to connect with you and follow along with Metaprop?
- 44:03
- Unidentified speaker (speaker_2): Uh, you go to our website, metaprop.com. Uh, you should listen, subscribe to my podcast. It's called Between Zach and Aaron with Zach Aarons, because uh I have a partner named Zach and a partner named Aaron, and my name is Zach Aaron. So we sort of took advantage of that, ripped off between two ferns a little bit. Um, if you've seen that, excellent. Uh, also by
- 44:23
- Unidentified speaker (speaker_2): Zach, Zach Galfinakis. Um
- 44:26
- Unidentified speaker (speaker_2): you can connect with us on LinkedIn uh
- 44:31
- Unidentified speaker (speaker_2): as well under the Metaprop handle, or you can come to one of the many community events that we host. For example, this week, week of October uh 19th, we're hosting New York City Real Estate Week. Uh, tons of different uh activities will be at all of them, CRE Tech, Commercial Observer, Power.
- 44:51
- Unidentified speaker (speaker_2): 100 event. Um,
- 44:54
- Unidentified speaker (UNKNOWN): so
- 44:54
- Unidentified speaker (speaker_2): we try to sort of continue to maintain an active community presence.
- 44:59
- Unidentified speaker (speaker_0): Love it. Staying busy in the community, hosting a podcast and joining us today. The one and only Zach Aaron's a Metaprop. Zach, thanks so much. I loved
- 45:08
- Unidentified speaker (UNKNOWN): it.
- 45:09
- Unidentified speaker (speaker_0): Thank you for the insights and congrats on all the success.
- 45:12
- Unidentified speaker (speaker_2): Thank you. Appreciate you having me on. thank
- 45:14
- Unidentified speaker (UNKNOWN): you.
- 45:18
- Unidentified speaker (speaker_0): All right, that'll wrap up today's interview. If you enjoyed the episode or a previous one, let the guest know about it. Share your thoughts on social or shoot them an email. Let them know what particularly resonated with you.
- 45:32
- Unidentified speaker (UNKNOWN): I
- 45:32
- Unidentified speaker (speaker_0): can't tell you how much I appreciate that some of the smartest folks in Venture are willing to take the time and share their insights. with us. If you feel the same, a compliment goes a long way. Okay, that's a wrap for today. Until next time, remember to overprepare, choose carefully, and invest confidently. Thanks so much for listening.