357 AI Agents Built Seven Startups
Welcome back to the deep dive. I have to say, when I first started going through the source documents
for today, I genuinely thought I'd found a typo. Oh, I know exactly which line you're talking about.
Yeah, I'm reading the executive summary and I see the capital raise target $52 million.
It's aggressive for a seed, but for a big series B, you see that kind of number.
It's in the ballpark at least. It's in the ballpark, but then my eyes drift over to the next column.
The projected valuation and there it is. There it is. A target valuation of 17 to 25 billion dollars.
Bullion with a B with a B in five years. And you know, my first thought is, okay,
this has to be some new AI chip company going up against Nvidia or maybe a quantum computing lab
in Silicon Valley or a massive pharma company with a miracle drug, something huge.
Exactly, but it's not. The documents we're looking at today, this whole package they call ATLA's V52.
It's from a group called the Black Hills Consortium and they're based in Custer, South Dakota.
Custer, South Dakota. I looked it up. Population is what around 2,100 people.
Yeah, give or take depending on the time of year. So not exactly a bustling tech hub.
Not at all. We're talking about a population that could comfortably fit inside a large high school
gymnasium, but they're putting forward a business plan with a valuation that rivals the GDP of a small
country. And that's not even the weirdest part. No, it's not because the business itself,
it's just it's all over the place. On paper, it looks like a mess. You've got bleeding edge AI
swarms. Right. Then you have a cannabis software company. Okay. Then a real estate holding company,
a media network with V tubers, government contracting software, and a cafe. A cafe. Don't forget
the cafe. I mean, it sounds like someone just threw a bunch of random business ideas at a wall.
Yeah. But the argument in these documents is that it's not random at all. Not in the slightest.
And that's what we're really going to unpack today. The Black Hills Consortium or BHC as they call it,
they argue they aren't building a single startup. Right. They use that phrase. They say they're
building an 11 entity economic development ecosystem. It's a mouthful, but the core idea, their entire
thesis boils down to one sentence. AI collapses the sales cycle. And not just the sales cycle. It seems
like it collapses the development cycle too. I mean, the claim that just stopped me in my tracks
was that they built seven operational startups in seven days. Seven days. The velocity is just
it's unheard of. How is that even possible? They call it an agent swarm. We'll get into the specifics,
but we're talking about 357 autonomous AI agents rating code in parallel. The output numbers are
in the docs, almost 400,000 lines of code. So this is where the whole death of the pitch deck idea
comes from. Exactly. Because why would you send an investor a static PDF with a chart you made
up in PowerPoint? That's the old way. Right. BHC is sending out a password and a live URL.
They're basically saying, don't look at our projections, log in and look at our live database,
look at our real-time financials. That's a power move. It's a David that says, we're not planning
to do this. We've already done it. The machine is on. The machine is running. So we have a huge
amount of ground to cover. This whole build for ever thesis. The agent swarms this wild plan
calls, settle the West. But I think we have to start with a money starts. Because an ecosystem is
just a concept without an engine. In this case, the engine is a company called GrowWise.
That's right. GrowWise, which is legally monumental highs LLC, is the designated cash cow.
It's the one entity designed from day one to be massively profitable and to fund everything
else in the flywheel. And when you say massively profitable, the projections are frankly
unbelievable. They are. The documents are targeting $394.5 million in year one. And that's just
from the software's service revenue. Okay, we have to pause there. Almost $400 million in the first
year. Even the legends of saws, you know, the slacks, the zooms, the snowflakes of the world,
they burned cash for years. They didn't hit that kind of revenue for what 5, 6, 7 years.
How could BHC possibly justify that number right out of the gate? They justify it by arguing
they're not really starting a company. They're stepping into a power vacuum. Yeah. To understand
the GrowWise projection, you have to understand the absolute state of chaos in the cannabis software
market right now. Right. I've heard bits and pieces. It's a mess. It's more than a mess. It's a
collapse. We call it the dutchy vacuum. And dutchy for our listeners is sort of the 800 pound gorilla
in that space, right? Yeah. Where they were. They were absolutely. They were the VC darling during
the whole green rush raised a ton of money, big valuation. But if you look at the data in the
Atelia's documents, and these are verifiable numbers, dutchy's valuation has collapsed by 89%.
Down to around 400 million from a high that was well into the billions. Another major player,
FlowHub, is down 88%. The marketing incumbents are, and I'm quoting the docs here,
bleeding out. So what's happening to them? Is it just a market correction? Or is there a fundamental
problem with their products? What's it like being a customer, a dispensary owner right now?
It's a fragmented nightmare. Imagine you run a dispensary. You're probably paying on average
somewhere between $3,000 and $8,500 a month. A month for software. Per month, per location.
And that's not for one clean, elegant solution. That's for a whole stack of five, six,
sometimes seven different vendors. So you have one piece of software for your point of sale,
your cash register. Correct. Then you have another one for your inventory management,
which is crucial. A third one for your customer loyalty program. A fourth for your e-commerce
menu. And most importantly, a fifth one just for compliance. And I'm going to guess none of them
talked to each other very well. Barely, if at all. So you have data silos everywhere. Your staff
is doing manual double entry of data, which, you know, and evidently leads to human error.
And when an error happens in cannabis, it's not a small problem. It can get your license pulled.
So it's expensive, it's inefficient, and it's risky. Exactly. But the real trap, the thing that
keeps dispensary owners locked into this broken system is the hardware. The hardware lock-in,
it's the classic strategy. It's the Apple strategy on steroids. To use dutchy or flow hub,
you can't just download the software onto any iPad. Oh no, you have to buy their proprietary
terminals, their specific scanners, their branded receipt printers. So a dispensary might have
spent, say, 50 or $60,000 just on hardware. Now, even if a better cheaper software comes along,
what do you do? To switch, you have to throw all that expensive hardware in a dumpster and start
over. It's a moat they've built, but it's a moat made of customer pain. That's a perfect way to
put it. And this is the door that grow wise plans to kick down. So they're coming in with a
fundamentally different model, a completely different approach. They call it an infrastructure protocol.
Grow wise isn't positioned as just another app. It's a layer that runs on top of anything.
Meaning you can keep your old hardware. You keep your hardware. You've got to beat up Android
tablet. Fine. You have an old Windows laptop in the back office. It'll run a brand new iPad pro.
Great. They're totally hardware agnostic. They're systematically eliminating every single ounce
of friction from the switching process. That alone is a huge deal. But I saw in the fintech section
of the documents that they're also competing aggressively on price, specifically on payment
processing. This is the wedge. This is how they get in. They're offering ACH transfer fees of
one to 1.5 percent, which is it's almost predatory pricing. It's so low. You have to understand
because cannabis is still federally illegal, payment processing is a high risk business. The
industry standard is somewhere between 2.5 and 3 percent. So grow wise comes in at 1.5.
You are literally giving that dispensary owner a 1 to 1.5 percent raise on their entire
gross revenue overnight. For a low margin retail business, that is not a small number.
That's the difference between being profitable and going out of business for some of these shops.
For many of them, yes, it's game changing. It's very similar to the playbook that the company
toast used to take over the restaurant industry. Right. Toast didn't just win because their
point of sale screen was prettier. They won because they handled the payments better and cheaper.
Precisely. And BHC is very explicit about this. The documents forecast that by the fifth year
of operation, $400 million of grow wise's annual revenue won't be from software subscriptions
at all. It'll be from payment processing. Purely from being the financial plumbing. They're not
just building a sauce company. They're building a fintech bank for an entire industry.
So the product is better. It's cheaper to run. It's easier to switch to. That'll make sense.
But even the best product in the world can fail if you can't get the word out.
Sales is a slow grinding process. You need an army of sales people making calls.
Usually, yes. And this is where that full market capture strategy comes in. And I'll be honest
reading this section. It felt, I don't know, a little terrifying. It feels relentless.
It's industrial scale sales. It's what happens when you apply AI swarm logic to a go-to-market
strategy. They're claiming they can reach every single licensed cannabis operator in the United
States within 90 days. Not a segment. Not a target market. All of them. Every single one.
How is that physically possible for a team of 51 people in South Dakota?
It's only possible because they're not starting from zero. They have the data. The ATLA
and the documents name refers to their database. It's a live super base instance that
as of the documents writing contains 18,786 verified cannabis accounts.
What does verified mean in this context? Is it just a name and an address?
Oh no, it's deep. It means they have the business owner's name, the store manager's direct line,
the state license number, the license expiration date, what software they're currently using,
the docs claim 98.9% phone number coverage and 95.6% email coverage for the entire US market.
So this isn't some list they bought from a shady data broker. They've essentially built a
complete intelligence map of their own industry. A surveillance map, you could call it.
And because they have that map, they don't need 500 people in a call center. They use the AI
agents and automation tools like NEN to orchestrate a simultaneous hyper personalized outreach campaign.
I think the phrase they use was spearfishing everyone at once. That's the one. So imagine
you're a dispensary owner. You don't get a generic email that says, hey, buy our software.
You get an email that says, hi, Sarah, I see your California retail license number C-C-O-0-0-0-0-0-1-2-3-L-I-C-A
is set to expire in 45 days. Based on public records, you're likely paying around $4,000 a month for
dutchy. With grow wise, your monthly bill would be 1900 and you'd save an additional 1.5% on your
payment processing. Here's a link to a pre-filled contract. And 18,000 other owners get a similarly
personalized message in the same week. In the same 90 day window, it's designed to create a
shockwave, a mass migration event. They're specifically timing this to coincide with the
sunsetting of dutchy plus, an older platform that's being retired in late 2026.
So thousands of businesses are going to be forced to migrate anyway. They have to pick a new provider.
And BHC's plan is to be the only logical choice waiting for them. They're positioning themselves
as the lifeboat right as the Titanic is going under. There is usually one big obstacle for a software
lifeboat in this industry, though. And that's compliance. We mentioned it briefly. M-E-T-R-C.
Ah, yes. M-E-T-R-C, the bane of every operator's existence. For those who don't know, this is the
government-mandated seed-to-sale tracking system right now. It is every single cannabis plant has
to be tracked from the moment it's a seed to the moment it's sold to a customer. If the numbers in
your store's software don't perfectly match the numbers in the government's M-E-T-R-C database,
at all times, you can lose your license. It's that serious. And the nightmare is that every state
has its own unique flavor of M-E-T-R-C. Exactly. California's M-E-T-R-C integration rules are
different from Colorado's, which are totally different from Oklahoma's. There are 19 different
state-specific integrations a software company has to buy and maintain. And building just one of
those usually takes a team of engineers months. The documents estimate it takes about 18 months of
dedicated engineering time for a traditional company to get certified in all the M-E-T-R-C states.
It's a huge barrier to entry. It's a regulatory mode that protects the incumbents.
But B-H-C claims they've already solved this. They claim to have automated it with the agent's
swarms. Because they've mapped out the entire industry, they've also mapped out the regulatory
schemas for every single state. They've essentially created pre-built compliance templates.
So they can just plug and play. Launching a new state instantly. Instant nationwide compliance
on day one. It's a monumental competitive advantage that allows them to scale in a way their
competitors simply can't. Okay, the business strategy seems incredibly thorough. Aggressive,
but thorough. I want to shift gears a bit to the how. Because we keep saying AI swarms. And I think
for a lot of people listening, AI still just means you go to chat GPT and type of question.
Right, and you get a nice paragraph of text back. Yeah, but that's not what's happening here.
You can't build a complex multi-state compliance engine by just chatting with an AI.
No, you absolutely can't. That approach will fail nine times out of 10. To do what B-H-C
is doing, you need a different kind of architecture. The secret sauce, as they describe it,
is something called MCP. MCP model context protocol. That's it. So think of a standard large
language model like Claude or GPT-4 as a brilliant brain floating in a jar. It knows a lot about
the world from its training, but it's completely inert. It can't do anything. It has no hands,
it can't touch anything in the real world. It's trapped inside the chat window. MCP is the technology
that gives that brain a set of bionic arms and legs. It's an open standard actually pioneered by
the AI company andthropic that allows the AI model to connect to and use external tools
and data sources. Okay, so give me a practical example. What is a tool for an AI? A great example
is a SQL database connector. So B-H-C has an AI agent they named Sage. When a dispensary manager
asks Sage, how many units of our blue dream vape cartridges do we have left in stock? Sage
doesn't just guess an answer based on what it read on the internet. Right. It uses the database
tool provided through MCP to execute a genuine SQL query on the dispensaries live inventory database.
It gets the real number say 142 and then reports that number back to the user. So it's grounding
its answers in real live data? Precisely. Or think about the marketing tool. The manager could say,
Sage, I want to send a 20% discount to every customer who bought edibles more than 30 days ago,
but hasn't been back since. And Sage can actually do that. Not just write the email, but send it.
It does the whole workflow. Sage uses the database tool to identify the right customer segment.
It uses a content generation tool to write the email copy and then it uses an email API tool
to actually trigger the send to that list of customers. It completely collapses the gap between
the user's intent and the final action. That explains how the end product works for the user.
But how did they use this to build the company so fast? That's the part that's hard to grasp.
The 357 agents. That's the swarm part. It's about parallelization. They didn't just have one AI
working on the project. They set up a hierarchy of agents, almost like a corporate or chart.
At the top, you have a lead agent. This one is running on the most powerful, most sophisticated model,
something like Claude Opus. Its job is to be the project manager. It takes a big task like
build the user login page and breaks it down into smaller sub tasks. And then it delegates those
tasks down. Not exactly. It delegates to dozens of sub agents. These are running on faster,
cheaper models like Claude Sonnet. And each one has a specialized job. One sub agent's only
job is to write the front end code in next.js. Another job is to write the backend API endpoint.
A third one writes the database schema. A fourth one writes the user documentation for that feature.
They all work at the same time. It's like a digital assembly line. It's a software factory.
And because these agents are AI, they don't need to sleep. They don't take coffee breaks. They
don't get bogged down in meetings. They just code 24 seven. That's the math behind their claim that
51 staff produce the output of 255. The humans aren't the ones on the assembly line anymore. They're
the factory managers, which leads to this new job title. I saw in the documents, which I found
fascinating MCP administrator. Yes. This is a really important part of their social and economic
mission. They're not trying to go out and hire 100 senior engineers from Google and pay them
half a million dollars a year, which would be impossible in Custer, South Dakota anyway.
Completely impossible. Instead, they're creating a new role. They are actively recruiting people
from the local service industry, bud tenders, baristas, retail clerks, and training them to become
MCP administrators. And the salary range they listed was what, $40,000 or $130,000 a year.
In a place like Custer, that's not just a good wage. That's a life altering income.
It's a path to the middle class that didn't exist there before. And the job itself isn't coding
in the way we traditionally think of it. So what are they doing day to day? They're supervising the
swarm. They're setting the high level goals for the lead agent, the reviewing the code that the
sub agents produce, looking for errors and providing feedback. They're the quality control.
So it's shifting the human role from being the carpenter who builds the chair to being the
foreman who directs the team of robot carpenters. That is the perfect analogy. And this all ties back
to their bigger mission of rural revitalization. They're trying to prove that you don't need a four
year computer science degree from Stanford to participate in the AI economy. You just need to
learn how to operate the new tools. This philosophy of showing your work of being transparent with
the tools and the data. It leads directly to that anti-pitch deck. We mentioned at the top of the
show. Growwise-jet.versel.app-board that URL. Yeah. I mean, if you're a venture capitalist,
you spend your entire life being well, let's say marketed to everyone rounds up. Everyone shows
the rosy as possible hockey stick chart. It's part of the game. But BHC is refusing to play that
game. They just give you a password and say, log in. Here's the raw data. See for yourself.
So with a big customer churns, the investor sees the monthly recurring revenue number drop
in real time. Instantly. If server uptime dips, they see it on the status page. If a new feature is
shipped, they can see the user adoption metrics. It's a level of radical transparency that's almost
unheard of. It sends a powerful signal. The signal is, we have nothing to hide because the machine
actually works. Exactly. In an industry, especially AI, that's just drowning in vaporware and fancy
demo videos showing a live functioning data rich dashboard is the ultimate flex. It's a whole
new level of operational maturity static content is dead. I saw that monster repeated a few times.
That's their belief. The PDF is dead. The slide deck is dead. The only thing that matters is the
live system. Okay. So we've got the money engine with grow wise. We've got the tech engine with the AI
swarms. But then the documents take a sharp turn into government. Suddenly, there's another AI
agent and his name is Theo. Theo. Yes. The sage of politics. He's named after Teddy Roosevelt. I
assume. Of course, the Mount Rushmore Black Hills connection. It all ties together. So what does
Theo do? Theo is an AI agent designed for a very specific user and overworked under-resourced
public servant in a small rural town. Think about the mayor or the county clerk in a town of 5,000
people. They're doing 10 jobs at once. Go the mayor, the head of sanitation, the grant writer,
everything. Right. And because of that, they are leaving millions of dollars on the table. There
are billions in federal grant money available for things like rural broadband infrastructure,
clean water projects, road improvements. But to apply for these grants, you have to write these
incredibly dense technical 100-page applications. And a small town can't afford to hire a full-time
six-figure grant writer. They can't. This is what BHC calls the grant writing bottleneck. And
Theo is designed to solve it. A town official can feed Theo the requirements for a federal grant.
Give it access to the town's local data. And Theo will write the entire grant application.
That's a huge value proposition. So BHC is planning to sell this software to local government.
This is the brilliant part. No, they're giving it away.
They're free. Completely free. But there is one small string attached.
There's always a string. To get your free license for the Theo software,
your municipality has to send its staff to the BHC grow campus in Custer, South Dakota
for a mandatory training and onboarding session. Ah, I see it now.
The software is the lure. The real business is the campus.
The training itself costs between five and fifteen thousand dollars per government cohort.
And while these government employees are in Custer for the week, where do you think they stay?
In BHC owned housing. The passcreek holdings properties. And where do they eat?
At the OP, the BHC owned cafe. It's another closed loop.
The software is the lead magnet that drives high margin revenue for the real estate
and hospitality parts of the ecosystem. It feeds the flywheel.
But there's an even deeper or more strategic layer to this. It's a political strategy.
A Trojan horse. Go on.
Think about it. BHC's main goal is to build a massive national cannabis company.
Cannabis is one of the most highly regulated industries in the country.
Who are the regulators? State and local government officials.
The exact same people who are now using your free Theo software.
The people who are grateful to you because Theo just helps them win a three million dollar grant
to build a new community center. So you're not some scary tech company coming to disrupt them.
You're their partner. You're their ally. You've already helped them. You've established trust.
So when the time comes for a grow wise to apply for a license in their state or when new
regulations are being discussed. The door is already wide open. You're a known, trusted entity.
It completely reframes the relationship. It turns potential adversaries into partners
before the conversation even begins. It's a political moat disguised as open source software.
That's that is next level strategic thinking. Most startups are just trying to get their next
customer. BHC is playing chess on a geopolitical board. Most startups are optimizing for short-term
growth. BHC is optimizing for long-term entrenchment. They want to become part of the civic and
economic fabric of the country, which brings us to the overall structure. The consortium
of 11 entities. I want to understand how they all fit together. The documents talk about a
war chest. The war chest is the central redistribution mechanism. It's the heart of the flywheel.
The most profitable entities in the consortium primarily grow wise are required to contribute 7%
of their gross revenue into the central pool. So every time a dispensary sells a product using
the grow wise platform, seven cents of every dollar in software fees BHC earns goes directly into
the war chest. Correct. And that pool of money is then used to fund the less profitable,
more developmental parts of the ecosystem. It funds oric labs, which is their internal startup
accelerator. And it funds outpost media. Okay, let's talk about outpost media because this is
another area where they flip the traditional business model on its head. Most companies see media
and marketing as a cost center. You have to spend money on Google ads or Facebook ads to acquire
a customer. It's often the biggest line item on a startup's budget. But BHC's philosophy is
everything is content. What does that mean in practice? It means they aren't hiring actors and
renting studios to make slick commercials. They're simply documenting the real work they're
already doing and turning that into their marketing material. They have I think nine different
channels and 32 planned shows. A show might just be a screen recording of an MCP administrator
managing a swarm to build a new feature. And I saw they're using VTubers virtual avatars as the
hosts. Yes, which is another clever cost saving measure. They have AI powered avatars named Sage
Greenleaf and Pahasapa who can host and narrate the content 24 hours a day. Because the raw material,
the screen recordings, the team meetings is a byproduct of their daily work. The marginal cost
of producing this content is effectively zero. But the value it generates is enormous. It drives
inbound interest. It drives them towards their ultimate goal, a customer acquisition cost,
or CAC of zero. If you can attract all your customers organically through content you produce
for free without ever paying Google or Facebook, your profit margins become astronomical. And you
control your own narrative completely. You own the platform. You are the media. Okay, so that's
the digital site. Then there's the physical side, the dirt, past creak holdings.
Boning the physical platform is just as important as owning the digital one. They have a 15 acre
property and custer they're developing into the grow campus. And the projected leap in valuation
for this real estate is just staggering. I saw that they listed as a $2 million asset today,
but they project it to be worth over $80 million by year 10. How does a piece of land in rural South
Dakota appreciate by 40 times? It appreciates because of the value of the activity happening on the land.
This is where the settle the West program comes into play. Right. This is their version of those
programs like Tulsa remote where cities pay tech workers to relocate. It's inspired by that. Yes.
BHC through their nonprofit arm, the Seed Foundation, is offering incentives to remote workers,
software engineers, designers, writers, AI specialists to move to one of nine partner towns in
the Black Hills region. So they're literally paying skilled workers to move there. They're importing
human capital. They're importing economic complexity. Think about the impact. If you convince 100
people who each make $100,000 a year to move into a tiny town of 2000, you have just injected 10
million dollars in new annual payroll into that local economy. And those hundred new people,
they all need a place to live, which BHC owns and rents out through passcree coldings. They need coffee
in the morning, which BHC sells at the OP. They need office space. They need services. BHC captures
a slice of all the value created by these new residents. The presence of that high skilled talent
then justifies the massive valuation of the real estate itself. It becomes a self-fulfilling
prophecy. It's like a modern digital age version of a company town, a very apt comparison. But
instead of the town being built around a coal mine or a steel mill, it's built around a server stack
in an AI swarm. And the final piece of that puzzle is Orick Labs. Orick Labs is the accelerator.
It's the mechanism for future growth. It takes the money from the war chest and invests it into
brand new startups. They take a 7% equity stake and they focus on three key areas, cannabis tech,
AI infrastructure and rural tech. So they're essentially breeding the next generation of companies that
will eventually plug back into their own ecosystem. They're creating their own deal flow and they can
offer these young startups something that no VC firm in Silicon Valley can. They can say,
don't just take our money. We'll also build your product for you. Come to Custer. We'll give you
access to our AI swarm and you'll have a minimum viable product ready to go in a week.
For a non-technical founder with a great idea, that is an incredibly compelling pitch.
It all just keeps looping back on itself. It's a dizzying vision. Which means we have to talk
about the numbers, the big scary numbers we started with, the exit or rather the lack of an exit.
Right. The build forever philosophy. This is maybe the most radical departure from the standard
tech startup model. Yeah. If you take money from a venture capital firm, a clock starts ticking.
You have five, maybe seven at most 10 years to either get acquired by Google or to have an IPO.
The investors need to get their money back, multiplied. But BHC explicitly states in their founding
documents, no exit. So how do they get anyone to invest? By modeling themselves after a very
different kind of institution, they cite the Will's door foundation. The Will's door foundation.
I'm not familiar. They own the watch company Rolex. I had no idea. Rolex was owned by a foundation.
Most people don't. And that's the point. Rolex has no external shareholders demanding short-term
profits. They don't have to worry about hitting quarterly earnings targets to place Wall Street.
They are free to plan in decades, not quarters. That's the source of their incredible stability
and long-term success. And BHC wants to create that same permanent structure for custer.
Exactly. Because they know what happens if they follow the typical VC path. They build this
amazing thing and then they sell it to a private equity firm in New York. And the first thing
the firm does is look at the balance sheet. And they say, why on earth do we have 150 employees
on a campus in South Dakota? We can hire engineers in India for a fraction of that. Shut it down.
The value gets extracted, the jobs leave, and the town is left with nothing but a hollowed-out shell.
So build forever is a defensive strategy. It's a promise to the community.
It's a legal and financial structure designed to ensure that the wealth generated by this AI
engine stays in the black hills for generations. But the investors still need a return on their
$52 million. If there's no sale, how do they ever get paid? Dividends. Massive recurring cash
dividends. Remember that fintech angle we talked about? By year five, they're projecting $1.65
billion in annual revenue. A huge chunk of that is high margin payment processing. The free
cash flow from an operation like that is astronomical. And that's what justifies the $17 to $25
billion valuation. That's the key to the whole financial model. A standard sauce company might
get valued at eight or ten times its revenue. But a fintech company, a payment processor like
Stripe or Toast, they trade at much higher multiples. 15, 20, sometimes 25 times revenue.
Because the revenue is seen as more stable and predictable. It's stickier. So by shifting the
business model from just selling software to becoming the financial railroad for the entire
cannabis industry, they unlock that much higher valuation multiple. It's the final piece of
the financial puzzle. It really does all connect. The cannabis software captures the businesses,
the businesses drive the payment volume. The payment volume justifies the fintech valuation.
The valuation attracts the investment to build the real estate. The real estate houses the talent
that improves the software. It is a perfectly architected, self-reinforcing flight wheel.
So as we start to pull back and wrap this up, what is the real story here? What's the big takeaway
for someone listening? Is this a story about the cannabis industry? I don't think so.
Not really. I think the big aha moment when you really dig into these documents is realizing
that GrowWise and the whole cannabis angle, it's just a Muguffin. A Muguffin, right? The thing in
a Hitchcock movie that everyone is chasing, but it's not actually what the movie is about. Well,
precisely, GrowWise is just the funding mechanism. It's a convenient, inefficient and fragmented
market that's right for disruption, which makes it the perfect cash register to get the flywheel
spinning. But the real product here is the model itself. The product is custer self-decoder.
The product is a replicable playbook for systemic, rural, economic revitalization.
They aren't just trying to build a billion-dollar company. They are trying to prove that high-tech
infrastructure and AI tools can be deployed to reverse the brain drain and save America's dying
small towns. It's that tension you see throughout all the documents. The contrast between the
absolute bleeding edge, tech next ages, 16, super base, agent swarms, and this rugged, rural,
almost frontier setting. Yeah, that tension is the core of their identity.
Their slogan is literally, settle the West. But this time, the pioneers are armed with code
instead of rifles. I just can't get over that one metric. 51 people producing the output
of 255. And that, I think, is the lesson for everyone, no matter what industry you're in.
What BHC is demonstrating is that AI isn't just another tool. It's a form of structural leverage.
It gives a small focus team the capabilities of a massive corporation.
They're betting that a small crew in the town of 2000 can outmaneuver and out execute loaded
billion-dollar incumbents. Yeah. Because they are native to this new way of working.
They're building a speedboat while their competitors are busy trying to plug leaks in their decaying
battleships. It really does leave us with a pretty provocative final thought. If a team of 51
people and 357 AI agents can build seven operational companies in seven days.
And if one small town in the middle of nowhere can use that leverage to build a 17-billion-dollar
ecosystem from scratch. Then what's the excuse for every other struggling town in America?
Is the old model of slow incremental growth just over? Is economic development now simply a
question of having the right data, the right AI architecture, and the will to execute?
BHC seems to think so. And if they pull this off, it doesn't just build a company.
It redraws the economic map of the country. I think you're right. Suddenly, your physical location
might not matter nearly as much as the quality of your digital infrastructure.
Well, I can tell you one thing. I'm going to be keeping a much closer eye on real estate prices
in Custer, South Dakota from now on. It might turn out to be the smartest tech investment you could make.
Thank you for walking us through this. It's been a truly mind-bending look at what happens when
the next wave of technology meets the American frontier. It was my pleasure. It's a fascinating
case study. And for everyone listening, thank you for joining us on this journey. We'll see you
on the next deep dive.