Wall Street Bought The AI Future | Ep. 29

Wall Street just made its move and it wasn’t subtle. In a single week, billions flowed into top AI labs while the largest financial institutions quietly pushed forward on tokenized real estate funds and blockchain trading rails. When those two waves hit at once, the message to fund operators is clear: the next decade’s winners won’t be the loudest voices debating whether AI is “real.” They’ll be the teams who install it, document it, and use it to move faster than everyone else. 

We walk through why coordinated mega-checks into AI aren’t a lottery ticket on hype, they’re a bet on AI adoption inside existing portfolio companies where productivity gains show up directly in EBITDA. Then we bring it down to street level: LP diligence is changing fast, and we’re already hearing the new questions that decide who gets allocated to. Can we use AI in underwriting to process more deals? Can we use it in investor relations to communicate consistently? Can we use it in the back office so operational cost per dollar raised keeps falling instead of rising? 

From there, we connect the “brains” to the “rails.” Tokenization lowers minimums, speeds up settlement, and introduces the kind of secondary liquidity most traditional funds don’t want to think about. We also flag the compliance reality: regulators are scrutinizing AI-driven investing claims, so marketing can’t run ahead of actual workflows. Finally, we switch to tactical mode with five practical AI workflows and a simple tool stack we can install quickly to level up prospecting, pre-call prep, meeting follow-up, IR drafting, and underwriting review. 

If you want to stay competitive as AI adoption and tokenized finance reshape fundraising, operations, and investor expectations, hit subscribe, share this with a fund manager friend, and leave a quick review so we can keep leveling the show up. What’s the first AI workflow you’re installing this week?

Get the Free Capital Raising Course
Get Fund Flow Os
Join the Free Capital Raising Community
Apply to the Founder's Circle

Follow me on instagram
Connect with me on Linkedin

Transcript

Wall Street Buys Brains And Rails

SPEAKER_00 0:03

1.5 billion dollars for the brains, $4 trillion for the rails, and in one week, Wall Street just bought both layers of the next decade. And every operator I know is still in Facebook groups arguing about whether AI is real or not. What the heck are they doing? Goldman Sachs, Blackstone, and Hellman and Friedman just bought $300 million each into Anthropic. This same week, OpenAI raised another $4 billion, $5.5 billion in seven days, all pointed to one thing putting AI inside the companies on Wall Street's books. And while the AI labs are getting capitalized, BlackRock, Franklin Templeton, and JP Morgan are quietly launching tokenized real estate funds. The New York Stock Exchange is building blockchain rails to trade them 24-7. Your $50,000 minimum deal is about to be competing with a $5,000 tokenized Blackstone tranche on a phone app. Wall Street is not asking for permission about this either. They're absolutely destroying the table while you said it. Right now, I'm going to break down what just happened, why your LPs are about to ask you a question you can't answer, and the five AI workflows that you can install by Friday for under $250 a month. Now stick around because this one actually matters.

Fundraising Mission And Sponsor Tooling

SPEAKER_00 1:22

Welcome to Funds on Fire, the podcast that ignites the passion of investment funds in capital raising. Here we turn the complexities of fund management into clear, actionable steps that drive results. I've invested into diverse real estate across the United States and manage thriving funds. And I'm committed to transforming lives through the vehicle of investment funds and helping others to do the same. Join me as we document the journey of scaling businesses, raising capital, and impacting tens of thousands of people around the world. My name is Devin Robinson, and welcome to Funds on Fire. Real quick before we dive in though, this episode is brought to you by Funflow OS. Funflow OS is the autonomous operating system for real estate private capital, not just another CRM you have to babysit. We're talking about FlowAI, an actual AI employee that builds your daily action list, runs your investor follow-ups, drafts messages in your voice, and keeps you compliant inside of your 506B and 506C with full evidence trails. It's the employee you can't afford yet. You can go to funflowos.com and get use the code FIRE to get 50% off your first three months, or it's free to start. Go to funflowos.com. Now let's get back into it.

Why Top Firms Back Anthropic

SPEAKER_00 2:30

Okay, story one, the brains. That's what we're gonna dive into first. Quick context. This is not like the Monday news hour that we're gonna start doing. This is like the midweek into week drop. There is one beat this week that is too big to wait until Monday and stuff into a five-story rundown. So we're gonna do it one off. All right, is that cool with you guys? Here's exactly what happened on May 5th. Anthropic announced a $1.5 billion joint venture with Goldman Sachs Asset Management, Blackstone, and Hellman and Freeman. Each firm wrote a $300 million check in the same week, same dollar amount, same vehicle, same week. Stop and think about those three names and who they are for just a minute. Goldman, Blackstone, and Hellman and Freeman are three of the most rigorous capital allocators on the planet. Committees stacked on committees, on committees, on committees, full-time investor relations teams, underwriting models that make most operators' diligence look like a coffee chat with vibes-based due diligence processes. They don't just write $300 million checks on vibes, and then they especially all don't agree on the same number, the same vehicle. And the same week, unless something is happening. The day before that announcement, OpenAI raised another $4 billion. Two AI labs, five and a half billion dollars of fresh capital in seven days. Imagine being on a panel at a real estate conference last quarter where the monitor where the moderator asked, Hey, is AI overhyped? And one of the panelists said, Uh, I think we're at the peak of the hype cycle. This is gonna fizzle out. Imagine being that guy this week, reading that headline, five and a half billion dollars of professional money called you wrong in seven days. Just brutal. Absolutely. Now, at the top of an anthropic release Opus 4.7 in April, their release notes described it as a step change for reasoning and coding. And OpenAI did the same thing. Now, now I have to read enough press releases in my life to know that every anthropic release blog says they all say the same three things. We have made improvements to model performance. We are excited about the future of AI. Reach out to our enterprise team to learn more. Translation, we do not actually want to tell you what we shipped because if we tell you what we shipped, you'll figure out which department of yours is going to get automated. The reality from the conversations I'm I'm having with these funds that I work with is that hedge funds are quietly rewriting their research desk all around Opus 4.7 or Chat GBT 5.5. I had a call on Tuesday with a fund manager up in Connecticut. He runs a long, short equity book. And he told me just casually, like he was telling me what he had for lunch, that he replaced two analysts last month with one cloud workflow. The two analysts were each making $180,000 a year. The cloud workflow runs on $20 a month plus API credits. He kept the senior analyst, fired the two juniors, gave the senior a raise to manage the cloud workflows, and his desk coverage tripled. $360, 80 grand a year gone. Replaced by an AI subscription that costs less than a coffee budget. And the desk is producing more research, not even less. That is not a press release, right? This is exactly what's happening on these calls with these people I'm talking

The Real Bet Is AI Adoption

SPEAKER_00 5:44

to. Now, pause. The question I want you to actually sit with is why do three competing private capital firms coordinate $300 million each into one AI lab the exact same week? The answer is the part that nobody on Wall Street Twitter is saying out loud. They're not betting on AI. They're betting on AI adoption inside the portfolio companies they already own. Buying NVIDIA stock at the high is betting on AI. Funding the install team that puts Claude inside every Fortune 500 you already own is betting on adoption. The productivity gain shows up directly on Ebita inside companies with contracts they already have a higher rate of return than buying chip exposure. So Goldman, Blackstone, and Hellman and Freeman are not playing the AI lottery. They're paying to install the AI faster than their competition can. Follow the chain one step further. This is the part that really matters to you. If those three firms are systematically deploying AI into the portfolio companies, then the LPs above them in the capital stack, the family offices, pensions, endowments, are starting to expect AI inside every fund that they back, including yours. Now I'm going to tell you a story that's real because I'm sure some of you are like, Devin, chill. This is just news. What does it actually mean for me? Okay, fine. Let me bring it

The New LP Diligence Questions

SPEAKER_00 7:06

home for you. I had a call last week with an LP. Family office has written checks into about 14 real estate funds over the last six years. Smart guy, calm cadence, chilling, reads the LP letters all the way to the bottom. He's the kind of guy you generally like talking to. Halfway through the call, we're talking, and he's like, Devin, I always ask, what is their AI strategy? I want you to sit with how surreal that sentence would have sounded 18 months ago, six months ago. That question doesn't even exist on diligence calls. 12 months ago, he would have laughed you out. He would have laughed. He would have laughed if you had asked him to ask it. 12 months ago, he would have laughed, he would have laughed if you even mentioned it. He would have said something like AI strategy, Devin, you're flipping houses. What are you what are you? Open AI now? Today, it is the standard diligence flow. After Anthropic's announcement on May 5th, the question is about to be in every diligence flow you walk into for the next two years. He wasn't asking if I had a cap rate forecasting model. He was asking three things, and most operators are missing all three. Question one, do you use AI in underwriting? Translation, can you process more deals than the next operator? Question two, do you use AI and investor relations? Translation, will you communicate? Will your communication be consistent across all of your LPs? Or am I going to be the one who gets the late email this month because Devin happened to be tired on a Wednesday, right? Like, do you question number three? Do you use AI in your back office? Translation, is your operational cost coming down per dollar raised, or is your fund a 2019 stack pretending that it's in 2026? Now, imagine being the LP walking into a diligence meeting, and the GP across the table says, I personally email every single person. The LP smiles, the LP nods. The LP also doesn't write the check because the LP just heard I can't scale past 40 investors, and my back office is going to break the second we cross 50 investors. The press release version of we're looking into it is a clean death sentence in 2026. And it's the AI equivalent of saying we have a really exciting roadmap. When an investor asks for your traction, we all know what that actually means. The good answer sounds like this Cloud reviews every deal memo against our underwriting model in about five to ten minutes. Flow AI drafts every LP update in my voice and flags 506B language risks before we even hit send. Bookkeeping plus reporting dropped from 80,000 a year to under 20,000 a year in 18 months. The LP is grading you on now. And here's the part that is going to sting a little bit. The funds answering those three questions clearly are the funds that just got allocated to. You didn't lose that LP because your deal was worse. You lost it because your operational stories sounded older. Man. Now, if you're sitting there going, Devin, I don't even know where to start with this AI piece. Good news. That's exactly what we do as a fractional chief AI officer. Most operations can't afford the $60,000 to $150,000 a year higher. Actually, it's even $200,000 a year in some places to three, I think it was like three to $500. So we plugged it in. We audit your stack, we build your workflows, we connect them into your operations. The link is in the description if you guys want something useful and um if that's something you're

Tokenized Funds And 24-7 Markets

SPEAKER_00 10:33

interested in. Now, while the AI labs are getting capitalized, the other half of Wall Street was rebuilding these rails. And here's what that means, and here's what's happening. And most operators just kind of absolutely have no idea. BlackRock, Franklin Templeton, JP Morgan, all three have launched tokenized real estate funds, live products, checks moving through them right now, this minute. The New York Stock Exchange is building blockchain rails for 24-7 trading of tokenized securities. And Deloitte's institutional report projects the tokenized real estate market goes from roughly $300 billion in 2024 to $4 trillion by 2035, 13 times growth in 11 years. Now, you might be wondering, how does this connect back to AI? Now, watch this. The reason this push is happening now, not five years from now, is because AI just made the back office cost of running a tokenized fund absolutely collapse. You cannot run a tokenized real estate fund profitably with the 2018 op stack. The reporting requirements, the water wallet reconciliation, the secondary trade KYC, the compliance language that has to ship the second a token changes hands. All of that breaks a traditional back office. AI doesn't break. So AI made the Rails feasible. Wall Street saw the AI shift, then saw the tokenization shift and went, oh my gosh, both of these at once? We can rebuild both layers simultaneously. That's not a coincidence. This is the sequence in which it happens. So let me make this concrete. What does tokenization actually do to your fund? Three things. One, lower minimums. Tokenization fractionalizes the minimums that used to keep retail and small checks out of institutional pot product. Your $50,000 minimum deal is suddenly competing with a $5,000 tokenized Blackstone tranche. The accredited investor about to commit to your fund is now scrolling past your deck because they can put one-tenth the check into a name they've heard for over 30 years. Second thing, instant settlement. Blockchain rails settle in seconds, not T plus two. Your investor wires money on Monday, and the SPV doesn't actually fund until Wednesday. The Blackstone token settles before the wire confirmation hits your inbox. Third, secondary liquidity. This is the one operators do not want to think about. Your traditional LP locks for seven years, five, three years, whatever. The tokenization and the tokenized investor can exit. Liquidity becomes a feature your competitor is offering that you are not. So what does this mean for your fund in general? Your moat is shrinking, not gone, shrinking. There's still real value in being the operator who picks the deal, runs the property, and reports back honestly. That part is not getting tokenized at least this decade. The capital routing layer, on the other hand, though, is being rebuilt right now. Now, here's where I'm supposed to put on a tie and tell you to go build a tokenization strategy for your fund. I've seen the LinkedIn post this week. Every fund manager on LinkedIn is going to post the same think piece by Tuesday, and they're going to use AI to do it in some way, and they're going to keep their M-dashes and it's going to sound like slop and sound the same. They'll all say the same three things. Tokenization is here, or we're excited about the digital future, or reach out to learn more about your forward-thinking approach to blockchain assets. I'll save you the time. They don't have one. Their blockchain strategy is a notion doc. Their nephew started on a Tuesday and hasn't opened the document since then. The good news for the operators who pivot to build your own blockchain rail this week, there is none. That is a $50 million infrastructure project. Wall Street already won that race three years ago. You are not going to out blockchain JP Morgan from your laptop on a Tuesday afternoon, even if you vibe coded. I'm sorry, I'm rooting for you, but it's just not happening. You're not the Rails, you're not the brains. You are the operator. There's one more piece of news on the Rails side that nobody is connected to the AI side,

Tokenization Lowers Minimums And Adds Liquidity

SPEAKER_00 14:41

and you need to hear it because this all connects. The SEC's 2026 examination priorities document just flagged AI-driven investing claims as an anti-fraud risk. Translation: any operator marketing the AI angle as a return driver without the infrastructure to back it up is now in the examiner's first pass review pile. Let me be clear. Because I have an SEC attorney as a partner, and I'm not trying to put any of you out in a bind. If you do not have AI inside your operations, don't start putting AI-powered fund manager in your PPM and don't put AI augmented underwriting on your pitch deck cover. Don't say we use machine learning to identify alpha for you in webinars, unless you can demonstrate the workflow when an examiner shows up to actually validate that because they will show up and the SEC is reading the same news cycles you are. They saw the anthropic announcement. They saw OpenIAI raise $4 billion. They're now staffing exam teams specifically to test whether the AI claims you make in market match the AI that actually runs. The dumbest thing I've heard last month, and I've been on fun Twitter, was a fund manager telling me he was going to rebrand the fund as AI first with no actual AI inside of it. Bro, this is not a rebrand. The tokenization claims are next on that anti-fraud list, I promise you. So the play is not go market AI. The play is go install AI. Then if you want to talk about it later, you have something real to point to. The math just doesn't work. Your raise doesn't support a 60-person team. And but here is what changed in 2026. 18 months ago, you needed a junior analyst to write the underwriting memo. Today, Claude Opus 4.7 writes it in 10, 12 minutes off your model in the data room, often more accurately than any junior analyst would. And because the junior analyst was gonna stress eat at his desk and miss the third paragraph, anyways. Six months ago, you needed an IR manager to draft 200 personalized investor updates. Today, Flow AI drafts all 200 in your voice in under three minutes with compliance flag baked in. So you don't accidentally promise a return in writing inside of 506B update and then end up explaining to your SEC attorney why your inbox is a problem. Last year, you needed a paralegal to prep diligence packets. Today, you describe the deal to Claude, point it at a data room, use a certain skill, and the diligence package drafts itself. Then a human reviews, then it goes out hours instead of weeks. Last year, your CFO spent four days a quarter producing investor reports. Today, AI pulls from your books, drafts the narratives, formats the charts, and the CFO reviews it hours instead of days. For the first time in 30 years, a four-person fund can match the operational throughput of a 40-person fund. Four equals 40. That's not just some hype math. That's exactly what's happening right now. That's exactly what's happening right now inside the funds that I'm working with. And here's the part that nobody's gonna say on Wall Street Journal: this is the biggest opening for minorities and women fund managers since the SEC let you advertise. Only 1.4% of the $82 trillion in AUM is run by minorities and women. 1.4% of 82 trillion. That stat has been stuck for 30 years because the operational gap. The big funds had the teams. We didn't. AI just deleted that excuse. The funds that win in the next decade aren't the ones that have the biggest teams. They're the ones that have the sharpest stack, the best people who know how to use AI effectively. If somebody's been telling you for years that your team is just too small to scale, that excuse is dead. Your lived experience is your competitive advantage. Your speed of execution is your moat. AI does the work that headcount used to do. And here's the part most operators don't see the big funds know it too. That is exactly why Goldman, Blackstone, and Hellman and Freeman just paid $900 million to combine to lock up Anthropic's enterprise capacity. They're trying to install AI faster than you can install AI because they know if you adopt this stack faster than they do, you're coming for their LPs. Do not let them install this faster. Okay, enough analysis, tactical mode. Five AI workflows, three tools under 250 bucks a month total. Install this by the uh Friday, and let's get after it. Workflow one, AI prospect research. The tool, pick clay or Apollo, either works. What it does, you plug in your LP target

SEC Risk Around AI Marketing Claims

SPEAKER_00 20:36

list. The tool pulls everything publicly known about each prospect, LinkedIn activities, recent investment moves, current portfolio, conferences they attended, and they post about what they post about mutual connections to you. Some of these tools also pull SEC filings, news mentions, podcast appearances. I mean, it's cool to know your prospect was on a podcast last month talking about why he hates real estate funds and saves you a meeting. What's cool is Funflow can also do this and completely contextualize it around capital raising. The cost of these things, around 150 bucks a month or for the tier that you actually need, not enterprise, just a working tier. The results, before your next LP call, you know more about that LP than they know about you. You stop opening calls with tell me about yourself. You open with, hey, I saw you wrote a check into so-and-so's fund last quarter. What did you like about that thesis? That alone moves 20% of meetings to a follow-up. Workflow two, the pre-call AI brief. Same tool, different use. Before every LP call, you generate a one-page brief. Their portfolio, their thesis, their last three LinkedIn posts, open questions you should ask, the competing funds they've already backed. AI does this in two minutes. You read it on the way to the call. You walk in sharper than the last GP they talk to. Done. Workflow three, AI meeting and follow up automation. The tool you could pick Fathom, you could pick Otter, you could pick Pick granola. What it does connects to Zoom and Google Meets, transcribes every call, summarizes the key points, drafts the follow-up meeting. Some will draft talking points for the next call based on what came up in this one. Cost, I don't know, 20, 30 bucks a month per user. Super easy. Install it. 20 minutes. Click, click, click. Do it. Make it happen. Results, you stop dropping the ball on follow-ups. Every ALP call gets a sharp, specific recap email in their inbox 12 minutes after you hang up. While the conversation is still fresh in their head, the most consistent comment I get from investors is Devin, you're the only operator who follows up that fast, remembers everything. It's not me. That's the automation. The tool does the work. I just press send. Workflow four, AI investor relations drafting. This is where flow AI inside Funflow OS comes in. You need a system that holds your funds full context. BPM, investor list deals, compliance docs, communication history, and layers AI on top of that context to actually do the work. Not generic Chat GBT, but a co-pilot that knows your fund and knows your LPs. Inside Funflow, that is what Flow AI does. It holds your fund's context. Draft those updates in your voice, runs your daily action list, keeps you compliant under 506B, 506C with the evidence trail. Make sure you jump on that. Go get it. The code is fire, 50% off your first three months. Easy. And even if you don't use Funflow, the principle is still the same. You need one AI co-pilot that holds your funds full context context. Otherwise, you're just stitching tools together with duct tape, and the LPs are going to fill it on the call. Workflow 5, AI underwriting review. The tool, Claude, Chat GPT directly pointed at your model before any deal closes. Drop the offering memo, the rent roll, the title work, and your underwriting model into Claude. Ask it to find the assumption you're most likely to be wrong about. Ask it where the deal breaks. Ask it what the seller is hiding. Claude's not going to catch everything, but it's going to catch the line item that the analyst who didn't sleep last night was about to glaze over. I mean, I've caught a couple of underwriting errors in the last couple months doing this. Bookkeeping that costs an emerging manager $80,000 a year in 2024 runs for under $20,000 a year in 2026 with the right AI stack on top of it. Pick the three tools, block one afternoon this week, install. When the next LP asks about an AI strategy, almost laugh. Not because the question's funny, but because you've already answered it three months ago. For 30 years, the operational gap kept emerging manager stuck. AI just deleted it from operator with a small team to fund manager running 40 person throughput on a four-person headcount. This is the identity shift this moment is asking for. When Wall Street allocates $1.5 billion to a single AI lab in a single week with three of the most rigorous capital allocators and private capital writing checks at the same time, they're not telling you AI is coming. They're telling you AI has already arrived and you need to be ready for it. The funds that adapted to the last 12 months are pulling ahead right now. The funds that adapt to the next 12 months, catch up. The funds that wait until 2027, there is no good news for them. They get acquired or wound down. This is the cycle. Every infrastructure shift has worked this way. Internet, mobile, cloud, AI, same exact pattern. The only thing that changes is which operators are paying attention. You are listening to this episode right now, which means you're already in the front third of operators paying attention. And the next move is just install it. Let me know what you guys are going to do. Let me know if you guys have been thinking about this, what you guys are planning on. Send me a DM on Instagram, write a comment, write a note in the comments, whatever. I'd love to hear how you guys are using AI in your company. And I have a quick favor:

AI Makes Small Teams Competitive

SPEAKER_00 25:57

if you've been enjoying the show, there's one simple way you can support us, and it's by hitting the follow button or the subscribe button on whatever app you're listening to on. And I want to level this podcast up in every single way possible, bringing you more value, incredible content, and guests with new strategies. Following the show and leaving a quick review really goes a long way in helping us to grow and continue to deliver top tier talent. And it's the only free thing I'll ever ask you to do, and it makes a bigger impact than I can possibly put into words. So thank you for being a part of the journey, and I'll definitely catch you on the next episode. So, as I always like to say, to great success and greater impact. Peace.