The 2 businesses AI can never make obsolete

1. Key Themes

Communities, networks, and marketplaces are the durable moats in an AI world

Sam's opening thesis is that non-AI people can still build large businesses that compound alongside AI. He frames the shift as technology losing its status as a moat while trust-based networks keep theirs. The two core examples are NADA 20 Groups (peer groups for auto dealers) and MODA Communities (vetted dealer marketplaces on Facebook groups). Sam says: "Pieter Levels has been talking about like what moats are going to go away with AI, and previously technology was a moat. That's not a moat anymore, and I think communities, networks, and marketplaces are still a really good moat" [00:10:18]. He opens the episode with the same claim: "there's still a lot of money to be made by building businesses that will grow with AI, and they won't actually be AI. And what are those businesses? Communities" [00:00:00].

The simplest peer-group format is an old, durable, under-marketed business

NADA 20 Groups has run since the 1970s with almost no product complexity, and Sam's description is the template. "It's the most simple peer group ever. All you do is you bring a P&L statement of your business of your dealership, and you sit down with 20 others, and you go line for line, and you explain your profit and loss statement" [00:01:28]. The mechanics: "you pay something like six or eight grand a year to get in, and then they put you in groups of 24, and the groups are non-competitive, meaning people who aren't in your region, and you fly to the same city three times once a year" [00:01:28]. Sam guesses it does "13 to $15 million a year in revenue" and calls the marketing "horrible," which is the opening: "I would either copy this and just do it a little bit better and run ads cuz I don't think these guys are ever going to run ads" [00:03:13]. Shaan's reply, "Dog, it says for 100 years they've been running," only reinforces the staying power.

Vertical peer groups exist for nearly every profession, and a 2.0 version is available

Shaan points to White Coat Investor (doctors and dentists discussing finance) and says "basically every profession kind of has these," arguing media tooling has changed enough that a new-generation entrant could win: "there's a whole new set of media tools and a whole new set of doctors and dentists that you could reach and build like a 2.0 version of this" [00:04:40]. Sam's targeting rule is boring professions where "the average person makes $1 million a year or like in take-home income... millionaire next door type of energy" [00:03:36]. The cross-profession version is BNI, a local lead-sharing network that Sam says "just sold, I think they have 800,000 members" at "low digit, thousands of dollars a year" with rumors of a billion-dollar-plus acquisition [00:06:03].

Whales fund the entire economy of enthusiast businesses

Both the MODA example and the gaming discussion converge on extreme revenue concentration. The MODA founder told Sam: "Virtually everything we do for is for free, and the top 1% or even the top .001% pay to keep everything running" [00:15:38]. Shaan generalizes: "Candy Crush has 82 million monthly actives, and all the revenue comes from 4% of the players, so 96% of the players are just kind of the cost of doing business. It's just the cost of finding those whales" [00:13:23]. The design implication, from his Twitch experience: "you just needed to give them new ways to spend" [00:10:57]. Twitch's sub-gifting feature "dramatically drove like, you know, revenue up per channel because it just turns out that the whales were ready to spend more."

Charge the supply side for access to a vetted super-enthusiast audience

MODA's model flips the usual marketplace. Sam explains: "A dealer pays to, uh, post there, and in order to post there, you have to do a vet check" [00:10:18]. The dealers pay "1 to 5 to $10,000 a year" and go through an interview plus a peer vouch. The community is split by price tier (under $10K watches, over $10K, and a "superfan" tier above $100,000), which lets them segment whales by wallet size. Sam estimates "somewhere between 5 and $10 million a year in revenue" with "over a dozen full-time employees" and "600,000 members across like 20 or 25 different communities."

Gacha and mystery-box mechanics are a mega-trend that converts choice into revenue

Shaan describes "ripping" apps where a user opens a digital Pokémon pack, and if the pull is a dud the app offers a buyback. "The genius of the thing was the buyback. The buyback changed everything" [00:17:38]. The economics: "they got 50 bucks of revenue. They give 43 back. Seven is free. They never had to sell to give you the card. They just put the card right back in the deck" [00:17:38]. He ties it to a lineage from Japanese gacha games to Overwatch loot boxes to physical toys, and shares a toy-industry data point: a friend's company "took the same things and put it in a mystery box, and it was 7x the revenue" [00:21:03]. The key mechanic is that attaching real-money upside to gacha behavior makes it far stronger than cosmetic loot boxes.

Remove friction from an existing desire (the Ev Williams rule applied)

Shaan cites the Ev Williams principle for ripping apps: "you want to build a billion-dollar business, just find a universal desire, ideally one that's been around for a while, and remove friction" [00:16:27]. Collectors already bought and opened packs; the apps put the pack on the screen with instant buyback. The company he cites, Triumph, did a billion dollars in revenue "in like the last like last year."

Unglamorous utility apps can be quiet behemoths

Sam updates on Private Photo Vault, an app that stores sensitive photos behind a lock so users can delete them from the camera roll. It has about a million ratings, and the founder told Sam the revenue and headcount numbers (which Sam wouldn't share): "It's a behemoth" [00:24:11]. The founder is a US-based solo-ish operator in Vegas, not an offshore app factory, which Shaan notes is unusual for this category.

DeFi and prediction-market precedent: ideas that "die" early often come back at scale

Shaan draws a pattern: prediction markets like Augur were built in 2015-2017 and died, then "Polymarket and Kalshi broke out, you know, 8 years later, doing the exact same thing" [00:28:48]. He predicts the same for DeFi: "I think the same thing's going to happen with DeFi" and says "people are completely overlooking this, and it's going to come back with a vengeance."

2. Contrarian Perspectives

The best AI-era business may be a non-AI, offline-meets-online peer group run by a nonprofit-style incumbent's weaker clone

Most investors chase AI-native products. Sam argues a better, lower-risk bet is cloning a century-old format and executing marketing better. His evidence is that NADA 20 Groups lasts "forever" with a site that "looks horrible" and still, in his estimate, grosses $13-15 million a year. "I think that you could do this... and I think it would be a great, great, great business that could last for decades" [00:03:30].

Choice is overrated: removing it increased revenue 7x

Conventional retail wisdom says consumers want choice. Shaan's toy-industry friend showed data across categories that mystery boxes win: "everybody thinks they like choice, but here's what the data shows across category 1, 2, 3, 4, 5. Like over and over again, this overperforms" [00:21:48]. The same product, repackaged as a mystery box, produced "7x the revenue" [00:21:01].

Free users are not the customer; they are a sourcing cost

Shaan frames the long tail as overhead: "96% of the players are just kind of the cost of doing business. It's just the cost of finding those whales" [00:13:23]. This inverts the community-building instinct to maximize member happiness evenly. The implication is to design for the top 1% and treat the rest as the funnel and the ambient crowd that makes the whale's spending visible.

Building on Facebook groups is risky, but the network leverage may outweigh it

Shaan concedes the platform risk ("they change the rules quite often") yet endorses the model because "you get the benefits of plugging in to such a huge network, right? So, it kind of giveth and taketh away" [00:10:57]. MODA's multi-million-dollar business with a dozen employees runs on rented land, which most founders would rule out.

Crypto's "idiots and geniuses" split: only the high-conviction extremes remain

Shaan claims the crypto crowd has migrated to AI, leaving "the incredibly smart, high-conviction die-hards who deserve to get rich, and the idiots who still think like the Vietnam War's still happening" [00:26:11]. Read as a contrarian signal, the thinned-out field is where the mispriced opportunities (DeFi) tend to sit.

3. Companies Identified

NADA 20 Groups (National Automotive Dealers Association)

A peer-group program for auto dealers (and now adjacent owners like HVAC, marinas, boat repair) that started in the 1970s. Members pay roughly $6-8K a year, meet three times a year in groups of ~24 non-competitive peers, and walk through each other's P&Ls line by line. Mentioned as the archetype of a durable, under-marketed community business. Sam: "It's the most simple peer group ever. All you do is you bring a P&L statement of your business of your dealership, and you sit down with 20 others, and you go line for line" [00:01:28]. He guesses it does "13 to $15 million a year in revenue" [00:01:28].

MODA Communities

Runs about 23 Facebook-based communities for buying and selling watches, collectible cars, gold, and jewelry. Founded by a Sioux Falls, South Dakota operator in 2018, vetted dealers pay $1-10K a year to list. Sam: "he has over a dozen full-time employees, and they have 600,000 members across like 20 or 25 different communities" [00:08:57], with revenue he estimates at "somewhere between 5 and $10 million a year" [00:08:57]. Mentioned as proof that trust-gated marketplaces on rented platforms produce real businesses.

White Coat Investor

A community and media brand for doctors (and dentists) focused on personal finance, which "started as a blog" and now includes Facebook groups and media. Shaan: "It's a community of of doctors, maybe dentists... for them to talk about more of like their finance side of their life" [00:04:09]. Mentioned as a model for vertical peer communities that could get a 2.0 rebuild.

BNI (Business Network International)

A local lead-referral network that puts non-competing business owners (realtor, garage door, lawyer, landscaper) in weekly meetings. Sam: "I think they have 800,000 members, and it costs low digit, thousands of dollars a year to join" and "they just got acquired... I heard rumors it was in the billion-dollar plus number" [00:06:31]. Mentioned as a validated, large-exit community model.

Hampton

Sam's membership community for founders, which Shaan calls the "peer group" Sam is running. Sam's plan: "what we'll eventually do is buy other communities, cuz I think we know how to operate them well" with a goal to reach "hundreds of millions a year in profit" and "billions" in revenue [00:07:01].

Club LTV

Shaan's prior community for e-commerce owners, referenced by Sam as the format he'd replicate in a less competitive vertical [00:03:30].

Triumph (ripping app)

A digital Pokémon-pack app with instant buyback that Shaan says did "a billion dollars in revenue" in the last year, founded by a former quant at Citadel and Bridgewater. Mentioned because a nearly unknown app reached a scale usually reserved for household-name social apps. Shaan: "an app you've never heard of, an app you've never heard of that's doing a billion dollars" [00:16:27].

Collector's Crypt

A blockchain-based pack-ripping and buyback product. Shaan: "They were doing something like 50 grand a month, and now it's 25 million a month now" [00:18:22], and later, "They do 100 million a month right now" [00:25:33].

Rip It (Logan Paul)

Logan Paul's entry in the ripping category, cited as evidence that the trend is a fad-scale wave [00:18:22].

Candy Crush

Cited as a canonical whale-driven game: "82 million monthly actives, and all the revenue comes from 4% of the players" [00:13:23].

Twitch

Shaan's former employer. He credits the gifted-subs feature for lifting per-channel revenue by giving whales a new way to spend: "it's literally like buying a bottle at the club" [00:10:57].

AppLovin

The largest advertising network for mobile games, where Shaan's Bebo co-founder was CTO. Mentioned as the source of insider data on what whales are worth and how games acquire them [00:13:23].

Private Photo Vault (Pic Safe)

An app for storing sensitive photos behind a lock, with about a million ratings. Sam met the founder at a Hampton event and said it is "a behemoth" in revenue and headcount [00:24:11].

Polymarket and Kalshi

Prediction-market companies cited as the late winners of an idea that failed earlier (Augur and others). Shaan uses them to predict a DeFi comeback [00:29:38].

Binance

The crypto exchange that, per Shaan, suffered a $5 billion hack and had a roughly $2 billion SAFU reserve funded from 10% of trading fees since 2018 [00:27:57]. Mentioned as an example of a rainy-day-fund design.

Augur

An early prediction-market protocol that died before Polymarket and Kalshi succeeded [00:29:38].

Tornado Cash

A crypto mixing protocol cited as a way to launder stolen funds [00:30:28].

Mercury

Sponsor; Shaan says he uses it across seven or eight businesses and has moved his personal banking there [00:23:04].

HubSpot

Sponsor; offers a free database of business ideas from the podcast and the Breeze Assistant for content creation [00:15:51].

Content is Profit

A new podcast hosted by Luis and Fonzi Camejo, promoted at the end [00:30:57].

Bitcoin / Overwatch / FIFA / FarmVille / Block Blast / Tinder / Snapchat

Passing references: Overwatch and FIFA as adopters of loot-box mechanics, FarmVille and Block Blast as whale-driven games, Tinder as the subject of a source-code analysis, and Snapchat as the revenue benchmark ("what, like 4 billion").

4. People Identified

Pieter Levels

Indie hacker and founder of several solo-run internet businesses. Sam credits him with the framing that technology is no longer a moat while communities, networks, and marketplaces still are: "Pieter Levels has been talking about like what moats are going to go away with AI" [00:10:18].

The MODA Communities founder (Sioux Falls, South Dakota)

Started a watch buyer/seller Facebook group in 2018 and grew it into 23+ communities with a dozen full-time employees. Mentioned for building a vetted-dealer model on rented land. His quote to Sam on whales: "Virtually everything we do for is for free, and the top 1% or even the top .001% pay to keep everything running" [00:15:38].

The Private Photo Vault founder

A Las Vegas-raised, non-offshore solo-style operator whose utility app has a million ratings. Sam: "just a guy, just a nice guy. He does live in Vegas, but he's a nerd" [00:25:12].

Triumph's founder

A former quant at Citadel and Bridgewater who built the billion-dollar ripping app around the buyback mechanic [00:17:23].

Logan Paul

Launched a competitor called Rip It in the ripping category [00:18:22].

Ev Williams

Twitter/Medium co-founder, cited for the principle: "you want to build a billion-dollar business, just find a a universal desire, ideally one that's been around for a while, and remove friction" [00:16:27].

The BNI CEO

"Really nice lady from Texas" whom Sam spent time with; she led BNI through its acquisition [00:06:03].

Shaan's Bebo co-founder (Firkan)

Former CTO of AppLovin, who gave Shaan insight into mobile-game whale economics [00:13:23].

The FarmVille founders

Shaan spoke with them about game economies, sources and sinks, hard and soft currency, and how to nurture whales: "it's literally like talking to a slot machine operator" [00:13:23].

The app-source-code analyst

A person Sam calls "that kid," who inspects apps' source code to explain things like Tinder's matching algorithm and game monetization walls [00:14:46].

Satoshi

Referenced as the Bitcoin creator whose dormant coins could cause panic if moved [00:26:33].

Luis and Fonzi Camejo

Hosts of Content is Profit, who built content teams for Red Bull and Orange Theory Fitness [00:30:57].

5. Operating Insights

Use an applicant-vetting gate to turn a free audience into a paid supply side

MODA's dealers must interview with the founder and then be vouched by an existing member before paying. The gate makes membership scarce and trusted, so dealers pay up to $10,000 a year for access to buyers. Sam: "you have to talk with him if you're a dealer, and you have to like do like a a vetting interview, and then another person in the community has to, uh, vet you, and then you pay 1 to 5 to $10,000 a year" [00:08:57]. Tactic: charge the side that is monetizing, and make vetting the product.

Create new ways for your top spenders to spend

The Twitch sub-gifting feature was not complex, but it let a whale buy 200 subs ("$1,000") in public view. Shaan's lesson: "It just turns out that the whales were ready to spend more. You just needed to give them new ways to spend" [00:10:57]. Audit your top 1% for friction, status, or visibility features that unlock higher ticket sizes.

Build a buyback or recycle loop so inventory resells and margin is free

In the ripping apps, a user who gets a dud is offered a partial refund, and the card goes back into the deck. Shaan: "they got 50 bucks of revenue. They give 43 back. Seven is free. They never had to sell to give you the card. They just put the card right back in the deck. It shuffled, they'll sell that same card again to somebody else" [00:17:38]. Look for products where the underlying asset is non-consumed and can be re-sold.

Test packaging that reduces buyer choice

Shaan's toy-industry friend A/B-tested open-shelf choice against mystery boxes and saw "7x the revenue" [00:21:01], backed by data "across category 1, 2, 3, 4, 5" [00:21:46]. Run the same test on any consumer product with collectible or variable-reward dynamics.

Segment communities by wallet size so whales have a room of their own

MODA splits watches into under $10K, over $10K, and a superfan tier above $100,000, so the highest-value members aren't diluted by lower-tier chatter [00:09:41]. Tiering by spend increases relevance, price discrimination, and dealer willingness to pay.

6. Overlooked Insights

The BNI acquisition rumor signals a roll-up market for community businesses

Sam mentions in passing that BNI "just got acquired" with "rumors it was in the billion-dollar plus number," and later says Hampton's long-term plan is to "buy other communities, cuz I think we know how to operate them well" [00:07:01]. Together these imply that community businesses have become acquirable assets at large multiples, and that a buyer with an operating playbook can systematically acquire underoptimized ones like NADA 20 Groups.

The same cohort of whales moves across products, so whale identity is portable

Shaan drops that "somebody who spends a bunch of money in Candy Crush [is] highly likely to spend a bunch of money in Block Blast. It's the same player and FarmVille" [00:13:23]. The implication is that the scarce asset is the whale list, not the game, which explains why ad networks like AppLovin are valuable and why a vetted whale community (like MODA's) is a durable asset that can host multiple categories.