SIP Live: Screensharing the Best AI & Startup Takes
Two operators screen-share viral tweets about AI and startups and rate each one sip or skip — human support as the new moat, agents that stay in their lane, and why a third space is worth the capital risk.
Posted
2 months ago
Duration
Format
Reaction
casual, opinionated, banter-heavy
Views
5.5K
126 likes
57 · 43
Big Idea
The argument in one line.
A livestream built around one simple mechanic — react live to a viral tweet with a single verdict, sip or skip — works because two operators with real P&L experience argue from lived business data instead of manufacturing a hot take for its own sake.
Who This Is For
Read if. Skip if.
READ IF YOU ARE…
A solo founder or indie hacker who wants fast, opinionated reactions to viral startup/AI takes instead of scrolling X for context themselves.
Someone weighing one general-purpose AI agent versus several narrow, specialized sub-agents for their small team.
A bootstrapper curious how experienced operators actually think about niching down an already-proven idea (baby food app, third-space venues, arcades).
Someone deciding whether to invest in 'systems' before their business actually needs them.
SKIP IF…
You want a structured tutorial with step-by-step build instructions — this is off-the-cuff reaction commentary, not a how-to.
You're looking for VC-scale case studies — both hosts explicitly favor small, low-drama, profit-first businesses.
TL;DR
The full version, fast.
SIP Live is Greg Isenberg and guest cohost Jonathan Courtney (AJ&Smart) reacting live to viral tweets about AI and startups, giving each one a 'sip' (agree) or 'skip' (disagree) verdict. The throughline across takes: AI is quietly shifting work from engineers to support teams and from single generalist agents to narrow specialized ones; human customer support is becoming a real marketing differentiator as competitors automate it away; declining 'third spaces' (bars, arcades, bowling alleys) are a genuine business opportunity but capital-intensive, best funded by a separate high-margin business; and founders keep confusing building internal systems with the actual job of marketing and selling the product. The practical conclusion: stay small, specialize narrowly, and don't build systems before you desperately need them.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
"Starting Soon" countdown card before the stream goes live.
10:04 – 12:12
02 · Cold open — sip vs. skip explained
Greg introduces the show's premise and guest cohost Jonathan Courtney, explains the sip (good take) / skip (cold take) chat-voting format.
12:12 – 17:30
03 · Take 1 — Baby food app making $1M/month
A tweet about Solid Starts, a baby-food guidance app making $1M/month; the hosts discuss price-as-trust-signal and niching the idea to pets or specific geographies.
17:30 – 24:56
04 · Take 2 — "Customer support is eating engineering"
Sahil Lavingia's tweet sparks a discussion of AI turning support teams into de facto engineers, plus human support as a marketing differentiator (Trade Republic, Basecamp, Eight Sleep, Amazon's decline).
24:56 – 28:07
05 · Take 3 — "Two pizzas is too much pizza"
David Pan's tweet on shrinking team sizes in the AI era; the hosts agree a solo builder plus one seller is now a viable team.
28:07 – 44:01
06 · Take 4 — Fewer places to relax and socialize
A graph showing US bars, bowling alleys, marinas, and movie theaters all declining per capita; long discussion of third spaces, capital-intensive physical businesses, VR arcades, premiumized arcades, and recreating a nostalgic college-cafe vibe.
44:01 – 46:09
07 · Take 5 — Jason Fried: the shutter-button metaphor
Jason Fried's tweet compares bragging about AI-shipped software to bragging about photos taken with the shutter held down; both hosts agree, tying it to CEOs who chase 'systems' instead of running the business.
46:09 – 51:27
08 · Take 6 — "Building a company in 2026 is two jobs" (first non-sip)
Jess's tweet argues every company is now a software company; Greg pushes back — his only outright disagreement of the show — arguing founders get lost in systems and forget marketing is the job.
51:27 – 54:05
09 · Take 7 — One agent or many specialized sub-agents?
Nick's tweet (Orgo) argues for narrow, role-based sub-agents each with their own computer to limit blast radius; both hosts agree, with Jonathan citing his own company running separate agents on separate machines.
54:05 – 1:03:07
10 · Take 8 — AI FOMO and the Facebook-ads analogy
InternetVin's tweet about the pace of AI change prompts a long riff on cycling in and out of AI FOMO, the 'permanent underclass' narrative, and comparing today's AI arbitrage to early Facebook ads circa 2016.
1:03:07 – 1:05:15
11 · Take 9 — PewDiePie's "rich life"
Jonathan's own tweet about PewDiePie's "Studied the Slinky for 52 Days Straight" video as a definition of being rich enough to do nothing productive.
1:05:15 – 1:11:37
12 · Take 10 — Marketing stunts as an anti-signal
Benji Taylor's tweet on marketing-stunt spend as a red flag; discussion covers the Friend AI necklace, Red Bull and Sony Bravia as outlier exceptions, and Slack's early press coverage as proof that even 'organic' growth has a promoter behind it.
1:11:37 – 1:23:13
13 · Audience Q&A and wrap-up
Chat questions on senior-focused business ideas in Canada and reviewing creator tools on a future episode; closing reflections on the first show.
1:23:13 – 1:24:53
14 · Sign-off
"SIP LIVE — See you next time" closing card.
Atomic Insights
Lines worth screenshotting.
A baby food app can hit $1M/month by using a high price as a trust signal for anxious new parents, and the same psychology transfers directly to pet food.
As AI makes small product changes easier to ship, 'customer support is eating engineering' — support teams increasingly make the fixes engineers used to own.
Publicly promising a human on the other end of support, not a bot, is becoming a real marketing differentiator as more products lean on AI-only support.
Support quality tends to erode once a company stops fearing churn — Amazon's early support versus its later automated support is the textbook case.
One builder paired with one person who can sell and pitch is now a viable full 'team' for shipping and monetizing software in the AI era.
Before hiring, ask whether a defined loop or agent could do the job instead of adding headcount.
US 'third places' — bars, bowling alleys, movie theaters — are shrinking per capita, and the same isolation pattern shows up across Europe, not just the US.
Physical, in-person venues are capital-intensive and low-margin, so fund them from a separate high-margin business rather than expecting them to be profitable alone.
Pre-selling memberships through an existing audience, before a physical space even opens, meaningfully de-risks the capital investment.
Bragging about how much code you shipped with AI is like bragging about photos taken with the shutter held down — volume isn't a business result.
A company can run almost entirely on chaos with no formal systems up to roughly $10M in revenue, as long as the founder is actively marketing and selling.
Systems are worth building only once something is breaking repeatedly — building them earlier is often wasted effort.
Narrow, single-purpose AI sub-agents contain failures to one small domain; one general agent juggling hundreds of skills breaks in ways that are hard to isolate.
Being the only business in a niche using a given AI tool right now creates a real, temporary advantage — the same arbitrage window early Facebook ads offered in 2016.
A large, polished marketing push before a product has traction is often a warning sign rather than a confidence signal, unless the company is already an established outlier.
Products for older adults often fail because they're marketed as 'for old people' — older adults don't see themselves that way.
Takeaway
Sip or skip: how two operators actually judge a hot take
WHAT TO LEARN
The most useful lessons here aren't about AI tools themselves — they're about which ideas earn a founder's time: human support as a moat, niching a proven idea, treating systems as a luxury, and remembering marketing is never optional.
03Take 1 — Baby food app making $1M/month
A high price can function as a trust signal in categories where the buyer feels high stakes and can't easily verify quality, like feeding a newborn.
A proven business at scale usually leaves room for someone else to succeed at a fraction of the size by serving one narrow niche within it.
The same underlying problem (trusted guidance during a scary first period) can be ported across categories, such as new pet owners.
04Take 2 — "Customer support is eating engineering"
When a support team can see dozens of pieces of customer feedback a day, turning that feedback into a working prototype the next day is now realistic without an engineering queue.
Publicly committing to human-only support is a differentiator precisely because so many companies are moving support to bots.
Support quality tends to erode as a company gets comfortable with its market position — treat that decline as a warning sign, not an inevitability.
05Take 3 — "Two pizzas is too much pizza"
A single builder can now ship a full piece of software alone, but turning that into a business still benefits from pairing with someone focused on selling and pitching.
Before adding headcount, ask whether a defined loop or agent could get the job done instead of a hire.
06Take 4 — Fewer places to relax and socialize
Declining third-space usage is a real opportunity, but physical venues are capital-intensive, so fund them from a separate high-margin business rather than expecting them to be profitable alone.
Selling memberships or building an audience before a physical space opens meaningfully de-risks the capital investment.
A 'premiumization' pattern that revived one declining category (arcades, movie theaters) is a repeatable model to look for in any other category that moved from public spaces into the home.
The clearest business ideas often come from recreating a specific personal memory rather than starting from market-size logic.
07Take 5 — Jason Fried: the shutter-button metaphor
The volume of AI-generated output is not itself a business result — revenue and retention are the only outputs that count.
It's easy to mistake being busy building systems for doing the actual job of running a company.
08Take 6 — "Building a company in 2026 is two jobs" (first non-sip)
A company can run almost entirely on chaos with no formal systems up to a meaningful revenue ceiling, as long as the founder is actively marketing and selling.
Systems become worth building only once a specific thing is breaking repeatedly — building them earlier is often wasted effort.
The founder's core job of getting people to know the product exists doesn't go away just because a system produces the product faster.
09Take 7 — One agent or many specialized sub-agents?
A single agent juggling many unrelated skills tends to break in ways that are hard to isolate; narrow, single-purpose agents contain failures to one small area.
Keep each specialized agent's domain genuinely small — a single recurring task, not an entire department — or you reintroduce the same fragility.
10Take 8 — AI FOMO and the Facebook-ads analogy
Being the only business in a niche using a given AI tool right now creates a real, temporary competitive advantage that closes as the tool becomes common knowledge.
You don't need to track every model release — checking in weekly on what's actually usable is enough for most operators.
It's fine to disengage from AI news for months at a time without materially falling behind, as long as you re-engage before a major shift.
12Take 10 — Marketing stunts as an anti-signal
A very large, polished marketing push before a product has traction is often a warning sign, not a confidence signal, unless the company is already an established outlier.
Marketing stunts work as fuel once you already have a working channel and retention, not as a substitute for finding that channel in the first place.
Every successful 'it just grew organically' company still had someone whose explicit job was to get the word out — there's no such thing as marketing-free growth.
13Audience Q&A and wrap-up
Products for underserved demographics often fail because they're marketed as being 'for' that demographic instead of using the same energetic tone used for younger audiences.
A profitable niche business doesn't need a mission statement or a path to millions — a small, well-served local market fully justifies itself on its own.
Glossary
Terms worth knowing.
Sip / Skip
The show's up-or-down verdict system: a 'sip' means the hosts agree with or like a take, a 'skip' means they disagree with or dismiss it.
Loop
A repeating automated cycle, often AI-driven, that takes an input (support tickets, meeting transcripts) and produces an action or output without a human doing that step manually each time.
Blast radius
How much of a system breaks when one part fails. Used to argue for giving each AI agent its own narrow job and computer, so one failure doesn't cascade into everything else.
Two-pizza team
Jeff Bezos's rule that a team should be small enough to feed with two pizzas, historically interpreted as roughly 8-12 people.
Third space
A social venue that is neither home nor work — a cafe, bar, bowling alley, or arcade — where people gather informally without planning.
Asset-light business
A business model that needs little physical infrastructure or inventory, favoring high margins over capital-heavy operations.
Resources
Things they pointed at.
13:15productSolid Starts: Baby's First Foods
18:34productEight Sleep
18:34productTrade Republic Berlin billboard campaign (human support, no AI)
19:37productBasecamp
31:18productCursor (David Pan, CTO)
51:27productOrgo
33:25productAlamo Drafthouse
37:40productFogle/Vogle cafe (Oslo and Tokyo)
1:08:26productFriend AI necklace
1:08:26linkSony Bravia 'bouncing balls' ad, San Francisco
1:10:33linkSlack's early Verge coverage
1:03:07channelPewDiePie — "Studied the Slinky for 52 Days Straight"
“If the product works well for babies, chances are it works well for pets. That's just the rule.”
instant, transferable niching rule delivered as a one-liner→ TikTok hook↗ Tweet quote
18:34
“If you have to deal with the support, no... I'd rather they stop developing the product and just leave it as it is and improve the support and get more humans in there.”
“The team size for making something like a piece of software right now can be as little as one.”
short, contrarian claim about solo builders→ TikTok hook↗ Tweet quote
51:27
“Up until about 10,000,000 revenue, your company can be pure chaos, have almost no systems, be falling apart, be held together by sticks and spit, and it will still work.”
counterintuitive, specific-number claim that pushes back on 'systems' advice→ newsletter pull-quote↗ Tweet quote
49:20
“Building a product basically means nothing if you're not able to market it.”
punchy thesis line, no setup needed→ newsletter pull-quote↗ Tweet quote
59:56
“If you're the only tire company in New York City using [an AI tool] this week, you're probably going to get some form of advantage for a while.”
concrete, memorable local-business example of AI arbitrage→ TikTok hook↗ Tweet quote
1:22:13
“Older adults don't see themselves as older adults.”
short, quotable insight that reframes a whole market→ IG reel cold open↗ Tweet quote
Topic Map
Where the conversation goes.
10:04 – 12:12sparseShow format — sip vs. skip explained
12:12 – 17:30steadyBaby food app making $1M/month
17:30 – 24:56denseCustomer support is eating engineering
24:56 – 28:07steadyTwo-pizza team shrinks in the AI era
28:07 – 44:01denseFewer places to relax and socialize
46:09 – 51:27denseBuilding a company in 2026 is two jobs
51:27 – 54:05steadyOne agent or many specialized sub-agents
54:05 – 1:03:07denseAI FOMO and the Facebook-ads analogy
1:03:07 – 1:05:15sparsePewDiePie's rich life
1:05:15 – 1:11:37denseMarketing stunts as an anti-signal
1:11:37 – 1:23:13steadyAudience Q&A and wrap-up
The Script
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
17px
And we are live.
Hey, everyone. What's up? Welcome
to SIP Live, a live show where we're reviewing hot takes from the timeline, giving you some ideas.
I'm your host, Greg Eisberg, and joining me is my guest host, Jonathan Courtney, a shrimp millionaire.
Hello. I'm a shrimp millionaire.
If you've ever wondered what a shrimp millionaire looks like, this is exactly what it looks like.
This guy. I'm kidding. I he's a, you know, he's a a founder.
AJ and Smart, is that what you say?
I just realized that I'm hearing myself in the YouTube version of this while I'm talking the whole time, and now I just saw solved it and figured it out.
I'm very excited for this for this, Greg. Show us what's
happening today. So what's happening today is I felt that, you know, there's a lot of shows on the Internet, live streams.
There's a lot of podcasts, but they don't really pick what I think are the most interesting takes. Now I'm not talking about news takes.
Okay? I'm talking about ideas, things that, you know, scratch my head the right way, and I want us to just bring them on to a show with a friend.
Welcome.
Hello. I'm that friend. You're my friend.
And I wanted to basically say, is this a good take? You know? So I'm hoping that the crowd, the chat, is either gonna say, is this take a sip?
Which to me is a good take, a hot take, or a skip? Which to me is a cold take. So throughout this entire livestream, I don't know how long this is gonna go, thirty minutes, forty five minutes, an hour, who knows?
I want the chat to come in and just tell us how it is. I got chat going right here. So if you're in the chat, give us give us some emojis, juice boxes, give us, you know, the death emoji.
You can throw in a shrimp emoji for your boy, Jonathan Courtney.
And I just wanted to, yeah, you know, just get right into it, pick up some takes, and we'll see where this goes. So producer, can you change the screen to the takes?
And we will get right into it. Alright. So I've got a bunch of takes, but the first take I wanted to to give and by the way, Jonathan has not seen any of this.
I saw this tweet. It says, there's a baby food app making a million dollars a month. I was searching for baby food apps for my newborn, and I realized the market is absolutely insane.
The more expensive the app, the more parents seem to trust it. So you can see here, there's an app. This data is by Sensor Tower.
Solid starts, baby's first food, making a million dollars a month. Now, Jonathan, I know you're a father.
I'm a father too. And a baby. You're a baby and a father, which is one of the most interesting combos known to man.
And my question to you is, what do you think about this idea? You like it? Do not like it?
Do you think there's opportunities to create more baby apps? What's your reaction? Wait
a second. So what does this app do? Is it recipes or something?
We have a look at what it actually is it a recipe app, or is it for buying baby food? I think so. I don't know about you.
When I when I had my baby,
when my wife had our baby Or you had the baby. Just When when I had the So my wife actually bought a course around the first 100 foods that your baby should be eating.
And I don't know. It's a few $100. And so I don't know if you can see on the screen, but this is the app.
And it basically says the first 100 foods that you should be buying and making for trusted by pediatric feeding experts, meal ideas, and then you can track food and reaction.
So I was just when I saw this, I was shocked. Like, a million dollars a month for this? Like, you know That is insane.
It's insane.
Well, you know okay. Here's the thing. I I have a child.
The first couple of months with your first child, it's so kind of scary and confusing, and especially when it comes to health stuff, you just kind of want someone to tell you this is the safe thing, this is the right thing.
Food is very confusing, and so for me, if there was a product that's just really clearly saying these are the things you can feed your baby in the first hundred days, and these things will not kill your baby, most likely.
I think that, but the fact that it is a, the thing that it says here, the fact that the more expensive versions of the apps actually do better and that parents trust them more, yeah, mean, similar with like, I guess weddings and stuff like that, there are these things where people want to spend a lot of money because they just, you get one shot, you know what I mean?
You don't get like loads of tries over and over again to mess this up. So if I'm looking for things for my child, the price is helping me figure out, okay, does this company actually take this seriously, or is this just an AI generated thing?
So I think for me, when I'm looking at something like this, I didn't buy this particular app, but when I'm looking at baby foods, for example, in the store, I am kinda scanning the shelves and looking at the most expensive one.
I'm like, that's probably the safest one, even if that's not necessarily true. Yeah.
So I think, like, the take is
it's a sip. It's a sip. It's a sip.
I have two reactions to to this little sippy right here.
One is if the product works well for babies, chances are it works well for pets.
That's just the rule. Yeah. I feel like that.
Right? So it's like, okay. You know, what is the equivalent of you get your puppy?
What what does your first hundred days look like? And I think you can bring this solid starts idea to puppies that I think would work really well.
So there's that. That's number one. Number two is if something is making a million dollars a month, there's always an opportunity to niche this down into Yeah.
You know, a niche. So that could be a geography.
It could be solid starts for Canadian babies. It could be, you know, you're feeding them I maple
mean, my daughter can't eat cashew nuts or pistachio nuts. Right. And so you figure that out the hard way very early, and so if there were, I mean, basically, look, if this is making a million a month and someone would niche down, and they're okay with not making a million a month and making rather like 10 ks a month or 30 ks a month with a product like this that's probably not crazy difficult to maintain, I think that could be a pretty good business, especially if you don't half asset.
I do think that this is something where parents do actually care about the quality, so if you were to niche down, you know, $10.15, 20 ks a month for a product that has a built in market of new babies being born, I think it's worth actually putting a lot of time and effort into making it a really solid, no pun intended, product.
I actually like that, I like the idea of niching down into like the little subcategories of general baby food, and yeah, but you wouldn't make a million a month, but not everyone needs to make that, right?
No, Definitely not.
So this is take number one. Ashley, let's go to our next take.
In the chat, I'm just curious. Give that a sip or a skip so we know if if you like if you like the takes. Okay.
Actually, there's a bunch of really good takes here. We'll go to Sahil's take.
Customer support is eating engineering. Do you know do you understand what he means by this? I don't.
I don't know what that means. So I understand what it means because, you know,
I Well, because you're a genius. Can I try to just guess what I think it means? Yeah.
Guess guess what it what what do you think he means by this? Customer support is eating engineering. Okay, so in the age where it's much easier to create little if you have a product, if you have a digital product, if you have an app and there needs to be a change made to it, it used to be basically impossible.
You'd have to collect all the information, relay that back to the engineering team, etcetera, etcetera, etcetera. That loop is really complicated, whereas now, and we see this in our company as well, the people who are directly connected with the customers are actually able to make a lot of the changes that engineers used to have to make.
And that's what I would guess it means, but maybe it doesn't. No. You're you're you're pretty much right.
So basically,
the whole idea is, in general, what you're seeing with AI is, you know, people are gonna have their own personal brains, personal second brains, but companies have their own second brains.
And you're seeing just companies, small companies, big companies, AI native companies basically put all their data, meeting notes, customer support, those things in MD files and places where LLMs could read it.
So what's happening is, you know, there's been a lot of talk about engineering loops ever since Peter from OpenClaw and who else was talking about?
I think Microsoft man? I think Karpathy Karpathy was also talking about loops.
And Right. They were they were talking about it with respect to just engineering, how you basically are gonna do slash goal in codex or whatever. And you can basically say, like build on the startup.
Build on the startup. Build on the startup. But I think what's actually more interesting about loops is less about just the engineering, but doing loops with customer support meeting engineerings meet meets engineering.
So what do I mean by that? I mean that if you're getting 50 customer support pieces of data every single day, they're telling you what they like and what they don't like about your product.
If you can create a loop that says, every day, look at those transcripts, and every day, prototype a feature that solves their problem, and give them a goal.
The goal could be, you know, DAU over MAU increases, retention increases, virality increases, revenue increases, that I think that that's the beginning of how people are Like, the future of startups is exactly what he's saying.
Customer customer support is eating engineering. That's what he means, and I think that's a huge sip.
Massive gulp even, a massive Like, it's a sip, So much of a sip that it kind of goes back your throat into your lung, and then you cough it out, and then you gulp it. I can I actually add something to this? So this is this is tangentially tangentially tangentially I'm Irish.
I can't say the word. It's somewhat related. Okay.
A lot of products if if I'm looking at a really interesting way to differentiate products that are in a market where everything is kinda similar, one of the things you can really do to stand out is by clarifying that your customer support is not AI and that you're going to be able to access a human twenty four seven.
Uh, a company in Berlin here, uh, Trade Republic actually, I don't know if they're German based or whatever. They did this entire billboard campaign all over Berlin about the fact that they do not use AI tools, and it's a human every time when you call them. By the way, not sure how true that is.
I had we'll see. Um, but if you also look at Basecamp, Basecamp's website is one of my favorite examples of that. They talk about the fact that it's humans, but not only that, they actually show the person who's in charge of the team talking on the website.
And for me, I'm like, okay, I want you to be using these AI tools to improve the product, but when I have a problem, I wanna talk to a goddamn human and not to some bot.
And so for me, I have my pet peeve, you know, especially if you buy an expensive product, I buy the, you know, the Eight Sleep, the mattress. We love them.
Everyone loves them. It's all good. If you have to deal with the support, no.
You know, it's not. I'd rather they stop developing the product and just leave it as it is and improve the support and get more humans in there.
I honestly think that companies can use that as a massive, massive marketing customer advantage. Just, hey.
By the way, you'll talk to a person when you call here,
and then the other thing is cool too. Yeah. I I totally agree.
I think you I think people are willing to pay more, like, premium premium products that say you're gonna speak to a human being, I think, is is a big
is is is a big trend, and I like I like that. Dude, can I sorry? Can I say one more thing?
Of course. Can I say one more thing? You know you know the trend of.
Right? You know the general idea. Everything gets worse.
Like, a company, once it kinda owns the market, they basically kinda make their product worse and worse and worse, and she they they cut corners because they know you have to use it. The perfect example of that is back in the day, Amazon used to have the best customer support of all time. You're, like, instantly talking with someone.
It's amazing. And then over time, they're like, we own the market, so now we're gonna let a dog be the person on the other side of the screen just hitting the thing. So is something people talk about a lot.
There was even this big advertising campaign, I think it was in Norway, about how this is a thing that they don't wanna have, whatever. Everyone talks about it.
It's a really big common topic on x. Someone needs to do an advertising campaign around how their product is getting reversed in shitificated, and that their one of their company's, like, mottos is that this is not what they're gonna do, I think it is one of the things that for me kills brands faster than than anything else is when you're a big fan of a product, big fan of a brand, and then they're like, well, we have to grow it at all costs.
Fair enough. But can you do it in a way that doesn't it? And then can you use that for your marketing?
Big if true.
You know what's sad about a lot of tech companies is a lot of tech companies have the worst customer support. Like, because they're tech first, they they they they take it to the extreme.
And I think, like, customer support has to have a human element.
I agree. So that's that's my that's where I'm at. I'm I'm skipping I'm skipping fully 100% AI as customer support, but I'm sipping heavily.
I'm gulping heavily this idea of human support. Any cock.
Yeah.
Alright. We're gonna go into the next take. We keep we keep it we keep it moving.
Let's do let's do this one. David Pan.
He says, rest in peace to the two pizza team. Jeff Bezos gave us an all time great metaphor.
Keep the team small enough to feed with two pizzas. It shaped how engineering orgs got built for twenty plus years. He was right about small teams, still is, but in the AI era, two pizzas is too much pizza.
What do you think?
Well, first of all, I live in Europe, so two pizzas is probably much smaller than two American pizzas, but I get the point, I get the point. We're talking about American pizzas.
Well, the team size for making something like a piece of software right now can be as little as one. What I will say is if you do actually want to be able to make money from that product, you probably need to if you're the builder, you probably need to pair yourself with someone who's good at pitching and selling, and so I do think one American pizza is probably enough for the modern AI powered team, so I'm gonna give it a sip, and I also just think it's a cute tweet, ex post, whatever it is, so I'm gonna sip it.
So you're double sipping.
Mhmm. You're double sipping. Yeah.
I mean, it's hard to disagree with this one. Right? It's like, I think when you're building a startup or any company, really, you have problems.
You you encumber you have problems. And we always kinda say, like, oh, if I just hired this one person, he he or she will solve this problem. Yeah.
Like, what I've started to rewire my brain to do is before hiring someone, I'm like, could I create a loop that can actually get this done? What are or I ask myself, what are the jobs to be done for the person that I think is going to save me? And then I start by actually creating agents to actually solve those tasks.
That's what that's what I've been doing. And I agree that, like, two two pizzas what he's basically saying, like, the two pizzas team is, like, probably 12 people.
12 people eat?
How big are these pizzas?
They're big. It's it's America. You know?
They're they're big pizzas. Is that? They're they're big pizzas.
There's pepperoni on the pizzas. There's Pizzas here like this size.
Yeah. Like, two two two, three slices, you're done.
Exactly. No. We're talking about XL pizzas.
We're talking giant giant giant pizzas. So, yeah, I think I think I think David, he's the CTO of Cursor.
He's right. I like it. And he got me he got me thinking about, like, a team can't be more than seven people now.
Six people.
I think I I mean, look, he's right.
I just the pizza size, very different, you know? We're thinking about different sizes here,
David. When when he's right, he's right. Okay.
I saw this graph by Signal, who's just an incredible follow on x. He says, there's fewer places to relax and socialize.
So look at these graphs. US is established for a 100,000 people. Bars and nightclubs, down.
Bowling alleys, down. Marinas, down. Movie theaters, down.
And it's first of all, it's just sad to see this, like, straight up sad. Right? Like so my my question to you is when you see this, what's your, like, societal reaction?
And then also, from a business perspective, is there opportunity to create nightclubs now?
Is there opportunity to create bowling alleys, third spaces, movie theaters? You tell me.
What's your reaction?
Well, I literally started a business based on essentially this data. So first of all, SIP to the max. I like, for the fact that I think this is a huge problem.
A huge problem and a huge opportunity. I think that in general, everyone's coming to the understanding that kind of loneliness and isolation and staying at home all day, doomscrolling is probably not good for your mental health.
However, it's also extremely comfortable. I'm sitting right now in a very large office space for my business.
I have absolutely there's no reason for me to have this space. Like, logically, everyone can work from home, the team is distributed, etcetera, etcetera, but for me, it is very important to sort of maintain some sort of third spaces outside of my home because I start to just get too comfortable and get too isolated, and I'm not naturally the most social person, so I actually just don't end up meeting people then and I get weird and I kind of come out of my apartment after two days and I'm like, you know, I don't know if you know this feeling, if you've been at home too long and haven't seen people and you like go to the grocery store and you can't really make eye contact with people, you're kind of bumping into people, your balance is all off, you don't know how to interact and someone asks you a question and you're like hey!
And you're like that one? You're like what did you say to me? Who are you?
And I, yeah, so this is really one of these trends that I think is very unfortunate.
I think it's happening not just because it's happening especially because of convenience, like COVID exploded this whole thing where now everything, know, whereas before you would go out and get groceries, now you go, whereas before you would, you know, you have to go out and meet people to date or maybe even bump into people and interact with people to grow your career, now it's like you go on LinkedIn and cry, and I think that there's a huge, I know that there's a huge opportunity for bringing people together in relaxed, socialized spaces.
I know it for a fact because one of my companies is literally building a roaming one of those, and yeah, I'm a huge fan of the topic. I want to solve it for myself, my friends, and also then make money from it.
I think it's probably my it's probably a topic that I spend more time thinking about than anything else. So, like, straw into the bag of whatever we're drinking and sip it till the whole bag goes inside the straw as well.
So
I see in chat, Riley the designer says, I'm curious, is there a similar trend in non US countries, or is this a US only phenomenon?
This is not a US only phenomenon? I can tell you, like, honestly, a lot of people that I know who used to be super social, we used to hang out, we used to do things, now I don't know.
There's another part of this I don't want to really get into. It's actually a bit too probably going to be people on X go a bit crazy, but there are basically the combination of COVID, Internet, certain drugs that you can now get more easy everywhere in in the world has created the perfect storm of I'm staying home all day and watching YouTube and actually make it the week and actually make it two weeks and actually why would I do anything else?
And that is very much across all of Europe as well.
We basically met in a bar slash nightclub. You know? So Yeah.
You know? This We would have never known each other if we we didn't know each other digitally. We only know know each other from meeting in real life.
Exactly.
Yeah. So I agree that well, first of all, from a society perspective, this is just you just know it's bad.
So we all agree there. From a business perspective, I think there's a lot of opportunity.
That being said, the problem with these physical businesses is it takes a lot of upfront money to actually very expensive. It's capital intensive.
And you are you and I are tech guys. You know? So we're we like asset light, high margin businesses, you know.
Don't get mad at us, Chad. Yeah. I'm looking at you, Chad.
Don't get mad at us. That's that's just I don't think it's a worthy thing to pursue, though, and I'm using money from my high margin businesses
to fund this business. Like, I ran something last year, it cost me $9,110,000 to do it, and we made $110,000 so it broke even, but I got to have the feeling of actually being able to do something about this problem, but then the results of running it is that I got loads more ideas from my money making businesses, so you can feed the results back into what you're doing that makes money, but yes, it's super expensive, It's not a it's not like a sexy business, but if you especially, there are people listen Greg, there are people listening to your show who have a lot of money, and if they would want to solve this problem, they can do it on a local level, if they and and break even with it, have a lot of fun, make some amazing connections.
It's worth it's a worthy thing to pursue.
Yeah. I also think there's there's, like, a way to do it that it isn't super, super expensive.
Like, that's the truth. Like, if we're gonna go the traditional path about o around opening a bowling alley, like, yeah, you probably need to go to go to a bank, get a loan, do it, and that's gonna stress you out. But there's ways that you can actually leverage digital first to fund a lot of what you're doing.
So if you have a x account or an Instagram account and you can presell and sell memberships before you open it, I saw that there's someone who opened up, like, a twenty four seven cafe or or something in New York City, and he was selling 50 when people do this.
Yeah. I love when people do this. Yeah.
So I think I I I'm I'm skipping the traditional way to actually open some of these businesses, but I'm sipping a la max the idea of use the Internet to and, like, use what you know about creating high margin asset light businesses of the Internet, using distribution of the Internet, using AI on the Internet to get you to to get those people into the physical locations in a business model that's a little smarter.
And you already came up with, you know, you have this, I can't remember the actual acronym, but you have this general concept about building bootstrap businesses where it's audience first. Yeah. And that's what I did for my for my in person business.
I spent all of the time building an audience for the topic of doing creative my topic is doing creative work like painting, in person building, all of that kind of stuff, and I built a very small audience for it, and that audience is just very high intent for buying a product or for coming to like a retreat or a physical location, but all of the pre work was done digitally, building an audience, essentially using your content system for myself.
So it it it is digital first, but then the actual experience is in person. So
I'm looking at the chat. By the way, I'm glued to the chat, so please, please, let's let's keep it firing. It's good to read the chat.
I'm where where is where is the chat? The chat is it exists. It exists.
It exists. Alright. Well, I don't need to see it.
I see Anton says, what are your thoughts on VR arcades?
Is that a big business opportunity?
I'm not interested in that, personally. I just don't it's cool, but it's just not something that does anything for me, personally.
It's cool, but it's not cool. I don't know, look, VR, you know, I am the type of person who will I will buy the Steam frame, if any of you guys know what that is.
I am a gaming enthusiast. I'm gonna buy all this stuff and try it out. No matter how good it is, I always find that it's not a very social tool.
The motion sickness, even when it's getting better and better, still feels not great to some people, and the thing just covering your vision, it's just not something that I'm personally, you know, if I was gonna go that direction, the way I go is like land parties with my friends in person.
I much prefer that approach, and that's something I do personally, VR arcades.
Arcades, though, very interesting.
There's no charming there are no charming arcades. There's just depressing arcades with like they look like shisha bars.
That's the only thing you get, but there's no charming, you know, nostalgic, enjoyable place to hang out at an arcade.
That doesn't exist anyway in Europe, and that's something I have been considering wasting money on creating.
Here's my take on that. So the same upmarket that movie theaters went.
So movie theaters were, like, dingy. You know, in the mid two thousands, I think a lot of movie theaters got really, like, dingy and gross and, like, kinda sucked at the same time when DVD players were getting really popular and people just started and Netflix and people just started watching at home.
But then what ended up happening is there were a bunch of new brands in movie theaters that were just a bit more upmarket. So all of a sudden, you're able to have these bigger seats.
All of a sudden, you're able to drink wine. All of a sudden, you had really good service. And you'd pay extra for it, but, you know, there are a bunch of brands like Alamo, Draft House, and and others that came in and just absolutely crushed it.
A lot of them were funded by private equity, I think. I think that same model could be applied to arcades.
Arcades Yeah. Were awesome for a while.
Literally, like, seventies, eighties, nineties. And I think you and then the two thousands, like, people started getting home devices, the same sort of thing.
Then they started getting didgy, a lot dingy, and there's a lot of them closing. So I think there's an opportunity to do, like, a premium arcade meets LAN party focused on bringing people together, high quality food, you know, drinks, things like that.
look at anything where the use of it moved from having to be in a dedicated place to being in your home, and think about, well, what would be the thing that would actually make people want to leave their home for it, like that it's an event, that it's something that feels good.
I actually have a secret. I have a secret thing that I use for coming up with business ideas, and I am hoping basically no one does anything with this piece of information, but I'm curious what you do with it, Greg, or I'm curious what you think about this.
When I'm thinking about a new business, one thing I think to think back to is when I was in college, you can also say high school, when I was in college, what were like some of the things that just that I feel extremely nostalgic for?
And one of the things that came to my mind is in a lot of college campuses, there's like a cafe or a place where everyone eats, and there's always just loads of couches and everyone hangs out, and just it's a hangout area. And for me, I used elements of this, just this vibe, this feeling of people not having to plan, not having to say oh I'm going to be here at this time.
People just end up in that central space and hang out. I look at this and think about how can I recreate elements of this as a business in the real world?
There's an example of it. It's not my business. I wish it was.
In, um, if anyone wants to just look at an example of how this can work in real life, there's a cafe in Oslo called Fogle, I think it's called Fogle, v o g l e. There's one in Japan as well, in Tokyo, and it just has that exact feeling, it has that exact vibe.
It's always full, all day, all night. It goes from being a cafe into a bar at night, and this is the place when I came back to Berlin, was like, if I, at some point, get stupid enough, I am going to recreate that exact feeling here somewhere in Berlin, and that's something which I'm always looking for, is like what's something that when you walk in you feel, oh man, this is just like that feeling, and think about is there a way to recreate that?
Is there is that something that's missing where you live? And also maybe you lose all your money trying it, but we'll see.
I love the idea. I just I don't I don't know. It's a question.
I don't know if it's full you're fully able to recreate that feeling. Or No. No.
No. Or or you're just gonna be chasing that hive for the rest of your life. Potentially, yes.
But I know what you mean. Like, I I was thinking about that too. Like, while you're talking, I was thinking about, like, the cafeteria, like, the school cafeteria.
Mhmm. And, like, you know, there's certain friends are playing cards and certain friends are eating, and then you know they're all there. And it it feels like, what's up, everyone?
Like, you know everyone and such a good feeling. Right? You're planning.
You gotta plan you're not planning it. You're just you're just there. I think I get pieces of that.
Like, there's some, you know, communities I'm a part of where I get pieces of that, but it's not it's not as good as it was with with school. JC, I wanna move on to the next 10. Taking the first one.
Oops. I wanna write it down. Here we go.
Bookmarks. So let's see what we can do here.
K. I'm gonna do a Jason Fried one because I know you're
Jason a
Fried, cofounder of thirty seven Signals, says, bragging about how much software you're shipping with AI is like holding down the shutter button and bragging about how many photos you took. I was gonna say, what do you think about this?
But then I just noticed you wrote, love this.
Oh, yeah. I already commented on it. Yeah.
Honestly, everything Jason Fried says, I it's always I've never I've never seen a take by Jason that I didn't wholeheartedly agree with. So yes.
Yeah. So I I think my take on this is there's so much mind share that people spend on building products with AI, but there's so little mind share that people spend on marketing software with AI.
Dude, this is my nightmare peeve.
Yeah. You know the amount of, do you know the amount? You probably do.
Do you know some very wealthy CEOs who have fallen into the, I'm building, I'm building, I'm making or I'm just I'm making endless systems, just all about the systems, and then you're like, oh cool, so like you know you've been doing this for like four or five months, like how's the revenue?
Oh yeah, it's not going so good right now because like, yeah, you're no longer running your business, you're just making weird things now. Like, you forgot your job as a CEO. I honestly think a lot of CEOs I know, I know everyone talks about AI psychosis, no, AI is making them forget what their job was, and now they're doing like, and I can't remember who said this, someone pointed this out, now instead of a CEO doing work that's like worth 5 k an hour, they're doing the work that's worth like $5 an hour.
Like, I'm gonna go get my AI tool to go through all of these transcripts and find the summary for the last 10 meetings I was in. Who cares? You're not what are gonna do with it?
It's yeah. It's it's a thing. It's a it's a very big sip
to the max. Sip to the max. Chat, we're looking at you.
Drop an emoji, juice box emoji if you sip it, death emoji if you don't, if you think it's a cold take.
Sip it. I agree. I think I think this is a trend that's only the beginning.
I think a lot of people are starting to realize this. So let's see.
I really like this one. So Jess says, building a company in 2026 is two jobs.
One, building the product. Two, building the system that builds the product.
Every company is now a software company. Even if the product isn't software, the company is.
What do you think about that? He told me my first non sip. Wow.
Yeah, I'm not I can't be sipping on that now, so obviously like you can read that any way you want and say yeah, makes sense, but I think it would give if I were thinking about entrepreneurs coming into the market, I think this would probably give the wrong impression and I think it would even make the situation worse that a lot of people are in, which is they believe that and I literally just posted before we went live.
If you look at what I literally just posted before we went live, you'll see the exact thing I mean. I think that company entrepreneurs are getting obsessed with systems to the point where they forget that the job of a CEO and an entrepreneur is to mark is to get people to know about your thing.
So I don't really
agree with this one. I think But isn't the counter to the isn't the counter to that you need to build a system around like, a marketing system, for example, around getting your product seen?
I wouldn't agree. I personally would say that you can up it depends where you wanna get to. Up until about 10,000,000 revenue, your company can be pure chaos, have almost no systems, be falling apart, be held together by sticks and spit, and it will still work, whereas if you have perfect systems and everything is, you know, you've got your perfect system for making content and everything, but you as the CEO are not necessarily focused or doing the right things when it comes to, you know, content creation or how you promote your company, then the systems are all there for basically nothing, so I'm more of a fan of building systems only when shit starts to break horribly.
I'm not a fan of making systems before you desperately need them. I'm very much a profit first, revenue first CEO. I wanna see I wanna see the whole thing fall apart first before I fix it.
So you what you're saying is systems
are a luxury
Yes. Basically. And attraction Yeah.
And get a COO to do like, if your business is, when your business is making so much money and everything's falling apart, hire a really great operator or even better, hire someone who can use these really amazing AI tools to build out your systems.
I do think that it's possible for company owners, small business owners can get completely lost in the system building, and as you know, Greg, because you talk about this all the time, building a product basically means nothing if you're not able to market it.
It would be like it would be like your show, except you don't put it on YouTube. You just, like, show it to me every week.
Wow. And you've got this amazing system to send it to me every week, and I'm like, oh, cool. That's great.
Okay. But you must admit that the right system slaps.
You get it? Get I think the right the right system absolutely slaps, and I love I love when all that stuff works.
I just wanna, like, encourage people to
not start there. Yeah. Well, you're you have a lot of people who agree with you.
I see Justin says in chat, yes. I get stuck building systems before putting anything out.
I think a lot of people yeah. I think a lot of people do get stuck, so I'll give I'll give it to you.
I'll give it to you. I do think that there's something interesting, though, about once you've built your built you know, you're building a product and you're building a company, having some systems in place so that you remain consistent, that you're Yeah.
Moving towards a goal. Yeah. You know, what I talked about with loops before, for example, like SEO loops, ad loops, things like that, you know, helps,
and so So those are actionable. Those loops are actually actionable things that have results. I think it also depends on how long you how long it takes to see the results.
The thing yours, the things you're talking about, you can get you can see whether it's working within the space of one or two weeks. I think where a lot of people fall into the hole is like, I'm gonna spend the next three months building a crazy system to get all my transcripts summarized for me, and then every morning when I'm on the toilet, I'll get a text message from my AI bot telling me what my top priority for the last month was, and it's like, yeah, you don't need that.
Alright. Let's let's move. I'm getting hot about that topic.
The person who posted it, I'm sure, is much smarter than me. I really don't wanna say negative things about anyone. It's just was something It's late, bro.
It's too late. You already did. You I agree.
It's good. Systems are good. Okay.
Let's let's do maybe three three more takes. My friend Nick says, I've recently changed my mind on this. I used to think that one agent is all you need.
In fact, I still believe one agent is all you need to start. But the minute you start getting into hundreds of skills and automations that need to run, it gets to a point where it's painfully obvious that you need specialized role based sub agents that own entire domains in your org, each with its own computer.
His company is called Orgo, so he's saying Orgo, so that you limit blast radius when things go wrong.
Managing one agent with a thousand different skills and automations is just a nightmare, and he's he's quote retreating Sahil, said one agent is all you need. So the question is, do you need one agent, or do you need a lot of agents?
Any thoughts?
So I will just say that I am absolutely not the expert here on this, and I don't have I I have, like if if we're looking at agents' automations running in the background, I think we have, like, three across the whole of my company. May maybe I'm underselling it. Maybe it's, like, three or four.
Definitely separate agents for us. The reason is when something, so for example, really simple example of this is we have sort of like a SEO agent doing a couple of things in the background and we're using the same agent to sort of deliver the reports around what's happening, and that breaks very regularly, and when it breaks, you're not just fixing the reporting, you're fixing so many things at the same time and everything gets muddled up.
So even though I'm definitely not someone who really understands the depth of this and I'm not the one running this in my company, We run them separately and actually have separate computers for three of them, so that's just what we're doing.
I couldn't agree more with Nick. I think that things break when you have one agent doing a lot of things, and I think you start seeing magic happen when you give these, as he calls it, specialized role based sub agents that own domains.
Now he says entire domain. So where I kinda, like, skip is the word entire. To me, the domains aren't huge domains.
They're very, very small little things. An example, YouTube thumbnail person.
You know? Um, that's like a pretty small, uh, if you run a marketing org, YouTube thumbnails is like a a painfully small, like, maybe point 1% of the dom you know, whole domain.
So I think that's that to me, as of now, as of doing this livestream, you need a lot.
You need it needs to be specialized role based. That being said, going back to what we talked about with systems and stuff like that and overoptimizing, don't over optimize.
You know? Start small.
I still don't you know the YouTube thumbnail thing and you just said it really throwaway? I still haven't figured out how to make an agent. Figure I can't I can't solve the YouTube thumbnail thing at all.
Please help me. Send me a video. I need to somebody fix this for me.
I'll help you. I'll help you. It's Thank you.
I got you. I got you. Maybe we'll do a live.
Okay. Let's do Hi. Let's do, um, a few more.
So my friend InternetVin says, so many things are shifting at the same time.
It's difficult to process and understand what to do, but at the same time, it's so inspiring. Only five likes. What do you what do you think of this?
Oh, my god.
I don't know how this is for you, but and for the the people listening. I kind of phase in and out of the whole AI thing.
Like, as you've seen with me, sometimes I'm locked in and super excited, and then I'll go for like six, seven, eight months where I don't even use one of the AI tools or barely even touch it, and then something will bring me back. And so for me, I went, I had a phase where I was like FOMO, and I was like, oh my god, there's a new model every week, and, you know, there's always something crazy going on, and now I think now that we've been in this sort of exponential curve of the stuff just getting faster and faster for like two years now, I've kind of calmed down a bit where I'm like, it's also fine if I for example, with Fable, I asked you, I text I sent a this is a perfect example.
I sent a text message to Greg two weeks ago and I was like, dude, what's Fable? Just before it got banned, I was like, what's this Fable thing? Should I actually care about it?
Any AI questions I ever have, I just messaged Greg. And he's like, dude, it's really good. Then it got banned, so I didn't even get to use it, so then I forgot about everything for a while.
Then it came back, and the last, like, couple of days I've been obsessed and having a lot of fun and, like, catching up, and I watched the OpenAI event last night, and I wanna see everything and I'm consuming, and honestly, the usual pattern for me is in about two weeks I won't care anymore, but it's probably less so for you because that's your job.
So for me, I'm just in and out and I barely I'm less FOMO and I'm less stressed about trying to trying to process it, and I'm just treating it like if it's important enough, it'll flow into my brain, and I'm still not noticing anyone actually being able to use the majority of people, like I said, are getting locked into these systems building loops and not being able to actually make revenue from these tools.
So it's not really affecting me yet, but I'm definitely at the point where I'm like, oh, wow, it's really this stuff is really good now.
So I'm inspired mostly.
To me, it's a spectrum. So on one end, you have people who basically don't know anything about AI, maybe use ChatGPT, but are basically, like, late adopters.
On the other end of the spectrum, you have the, I'll call them, the permanent underclass people. So the people who are afraid of becoming in the permanent underclass, which, by the way, whole separate rant.
I I don't even I I I don't believe in in that whole, you know, narrative that that people are saying around escape the permanent underclass.
But those people, I think, are, like, the most stressed because they're like, I don't wanna be in the permanent underclass. First of all, the word permanent is in there.
It's it's it's like Sounds like you're going to hell. It's like you're going it's even worse because it's like hell on earth.
You know? It's hell on earth. And it's like, I gotta escape that.
Right? That's like a prison sentence.
It's a lifelong prison sentence. I gotta escape that. I need to know that, you know, this model came out and that model came out, I need to use this for that and that for this.
But I think there's a happy medium around, like, weekly, I think it's important to understand what what happened at this point.
Because I think that the opportunity right now to use some of these tools and the arbitrage that exists also in using these tools to deliver value to customers, uh, and creating high margin businesses is just out of control right now.
I think it's out of control. And I think for a lot of people who and it's sort of one of the reasons why I wanted to do these livestreams is, like, I don't think there's enough live shows talking about how to actually do this.
Like, how do you actually like, what are the what do I need to pay attention to? What are the tools? Why do I need to care about them?
How should I use them? What is fluff? What is not fluff?
And so I think that, you know, to Vin's point, it's difficult to process and understand what to do is because there's a fire hose. There's an absolute fire hose of tools and updates.
Um, but the reality is there's only, like, two things you should care about per week.
You know? Yes. Like I think Yeah.
They should be listening to your podcast, obviously, once a week and watching this live show, the end. The end. But you know what?
The the there people say this is this has never happened before, blah blah blah. There's a perfect analogy for this moment that we're having right now, and it's Facebook ads. Facebook ads when we started doing Facebook ads at my company, I think it was 2016, we were already late when it came to the general world of Facebook ads, but we were essentially first when it came to our niche, and Facebook ads allowed us to essentially print money for years until the rest of our niche figured out what was going on, and so I look at this in the same way where it's probably not that you have to know you have to be an AI expert, you know, immediately as soon as a new model comes out, but just being somewhat informed about what's going on, being able to use the tools more than them just being a search for you, like a Google search, like being able to use them in a slightly more nuanced way.
For me, I'm viewing this as the same feeling as Google ads, sorry, Facebook ads, where right now this is actually quite cheap and the arbitrage you get from it, the ROI you can get from it is quite amazing, and I just can imagine that, you know, two years from now, not only will the tools that are really good be very expensive, but also everyone will be just doing the same thing.
I mean, an example of that is right now, if you're in a niche that's if you're the only person in your niche using AI tools for your SEO, you're going to win SEO.
Like, that's something I can 100% promise you from direct experience of my own businesses. If you're the only one who's, and by the way, you don't have to be a crazy expert, but if you're the only one who's doing it, and let's say you run a tire company or something, if you're the only tire company in New York City using Fable this week, you're probably going to get some form of advantage for a while, and I think that's true, but I don't, I also agree with you that the extreme fear mongering and permanent underclass stuff is just crazy, so I think it is a combination of there are advantages to be had, it's stupid to think there isn't and it would be ignorant to just say it's all bull crap, but you also can just stop for a couple of months and chill out or just keep up lightly each week by listening to Greg Eisenberg's podcast, he didn't actually tell me to say this, I'm just messing with him, and you can also be happy at the same time, you don't have to become some sort of demon obsessing about this stuff.
Don't be a demon.
Doobie. Did you sip that way or not?
I sipped it big time. Like, I think I think a lot of people are feeling this, so I feel I felt like it was important to say.
You know what I mean? So let's, uh, you wanna do one or two more?
Dude, I am chill. I can do a few more. How about this one?
This is how this is how rich this is how rich I wanna be. And then you have a This is this is by J Ice Cream, Jonathan Courtney.
He says, this is this, comma, this is how rich I wanna be. I studied the slinky for fifty two days straight.
Yeah. I saw this Yeah. Saw this on YouTube.
I saw this thumbnail. Studied the slinky for fifty two days straight, and I was I actually watched the video because I was like, what the hell is this?
And this is I'm sure everybody knows who this is, like, the most famous YouTuber of all time. And for me, when I saw this, in a joking way and a serious way, I was like, I want to be so rich that all I, or like this is exactly how I want to be, it doesn't need to be more than that, it doesn't need to less than that, that I could spend fifty two days learning the slinky and everything's still all financially good, so that's for me, that's my definition right there when people ask like what's your, you know, the guy who wrote, what's it, Remit Sethi, he's like, what's your rich life?
My rich life is studying the slinky for fifty two days. You That's it for know you've made it when you've studied the slinky for fifty two days straight, and you can do that. What's really cool about PewDiePie
is I'm pretty sure he, like, retired off YouTube. Right?
He was, like, the number one YouTuber, and then he's like, I'm moving to Japan, and I'm gonna start a family, and, like, I'm out.
He, like, got out, and then he occasionally vlogs now,
but Definitely, yeah. He he's definitely one of the it's like the opposite, I guess, of the American way of getting rich and famous, where you kinda keep going up and up and up and it never stops, whereas I think the because he's Swedish, he was living in The UK, the European way is kind of, you're just like, I'm kind of done now, and so now I'm just going to do what I would like to do, and I'm a big proponent.
I am a European, so obviously I just feel that way. Oh, you're European? I had no idea that you were European.
I thought you were from the Midwest.
Yes, exactly. That's what you thought. But, yeah, I I'm dude, I am the exact I I look at PewDiePie's videos these days, my fiance and I watch them, and we're like, yeah, that's that level of that level of wealth, cool.
He's a cool he's cool. I sip this vibe highly.
So respects PewDiePie. Respects
respect for respect is due. I gotta learn that slinky.
Yeah. Let's get that let's get that slinky money.
Benjie Taylor, who I think is now the head of design of X, he says, I generally view any company spending more time on marketing stunts than product as a huge anti signal.
This is actually why I really like this is that it's like the it brings the nuance to the conversation of you have to promote your product, the CEO has part of their job is you have to promote it, but marketing stunts I feel are they're one of these things which kind of fall into this other area of where usually people like hire a marketing agency separately, spend tons of money on it just to do some big thing that Yeah, the marketing stunts, I don't know if I would view it like an anti signal, but they are to me when I see a really big marketing push for a product that's really polished and really kind of almost corporatized, I do get suspicious that this thing is maybe not going to actually survive in the real market.
You know, sometimes it's just about, you've seen this loads of times, like sometimes it's just a startup buying a domain name for like $10,000,000 and that's the marketing stunt, and then I'm like, I don't know about this now. I'm a little bit, you
know? Thoughts on Red Bull, perhaps the only exception to the rule, but their marketing is also closely tied to the product offering, and the product is already massively successful.
Yeah. I mean, Red Bull is just its Red Bull is a true outlier, and often you just can't use outliers as examples.
Whatever happened with Friend AI necklace? Haven't seen anything about it for a while. Didn't he spend over 1,000,000 on marketing?
Okay, so that would be an example of a launch video that was very expensive, a domain name that was very expensive. I don't want to I really don't want to take a crap on people's businesses, it's hard to run a business, but I would personally say that when I see a big marketing stunt from a company that I do feel, unless the company's already very successful, for example, you know, Sony, they did this Sony Bravia ad in San Francisco where they dropped all the balls down the hill and it was like a really big moment on the internet.
Yeah, but it's Sony, you know, it's probably fine. It's fine for them to do that, but, yeah, like a new startup that's getting loads of buzz, a buzzy startup, I don't know, man.
I feel like the buzzy stuff almost always just doesn't survive, you know what I mean? When you're going zero to one,
and what I mean by zero to one, I mean zero to some level of product market fit Yeah. The stunt isn't the thing that's going to get you there realistically. 100%.
The stunt is something that is fuel to the fire. Like, once you have something going and you you have the luxury of being able to experiment a lot more, yes.
Okay. Fine. Do some marketing stunts.
But if you're trying to get, you know, one or two channels working on to and and and get people to your product and and see if they retain and see if they share, you don't need a marketing stunt.
You really don't.
And I just this is I know it's hard to be nuanced on the internet, but a lot of people will look at this and they'll say, yes, marketing is bad, you should just focus on the product.
This is not, I believe also, what Benjie is trying to say. If you think about some of the most successful products on the planet, look at something where you would think, oh, there was no marketing, there was no Like Facebook or something, it all happened in the background and Mark is also very awkward, he wouldn't have been out there.
Or even Sam Altman with OpenAI, you think, oh yeah, well they were just building, it was a lab. These people running these companies are crazy good at selling what they have, promoting what they have, and getting more investors and getting people excited about stuff.
That's what their job is. They're going to the VC funds. They're going and collecting the money by pitching, by the way.
You just don't happen to see these pitches posted on Twitter or whatever, and so I still need people to understand that even when a product does seem like it kind of organically grew and it was just the product itself that made headways and it wasn't anything to do with no marketing happened, marketing is just a very kind of general word for that there was probably someone in that company whose job was to let other people know that this thing existed or else no one would have ever known it existed.
Like Slack, do you remember Slack in the very beginning? Like, I found out about Slack because I was reading The Verge, and there was a huge piece on it.
There was a huge piece about Slack on The Verge. I was at a design conference in New York. The CEO of Slack was on stage talking about making this product, and it's like, yeah, that's what this person's job is.
They're the promoter of the business. It's not just that these products just click click and now they explode. Most of the time, there's still someone being the promoter, but marketing stunts, anti SIP.
But what this guy said, yes, SIP. Yes, SIP. I agree.
I also I really like this reply by Morgan Linton.
I see high signal as companies hiring great people. I feel like that's, like, the greatest signal, So I do agree with that.
When I see someone who I really respect I mean, Benjie's an example of that. Like, he he's a friend of mine.
And when I saw he's gonna x, I was like, woah. That's, like, that's really, really cool. You know, he's a founder, and he's gonna join a company.
So agreed, agreed on this.
I definitely agree. I think I agree and I just hope that people understand the nuance of it because, you know, the thing, I think that Silicon Valley view is that marketing bad. Marketing bad, building product good, but they are basically teaching a lie because their founders spend all day on podcasts promoting their companies.
You know, they're like, yeah, well, you know, okay, Elon's a builder.
I don't think I've ever seen a podcast without Elon Musk in it, know what I mean? His job as part of his company is also to promote his businesses, and that's part of that's part of the role and the the job of being a CEO.
People are probably sick of me saying this. I'm gonna stop repeating myself now. I mean, I'm not.
I'm not. I like it. Your job promoter.
What do you think?
Me? Yeah. What'd you think of the first ever SIP live?
Chat, what did you think of the first ever SIP live? Not sure how long how many of these we're gonna do, but having fun while doing it. I I certainly enjoyed myself.
What did you what did you think, JC? I really enjoyed it. I mean, I'm complete
Greg sees the chat, I think. I'm completely isolated just looking at Greg on the screen here. As you can see, I'm not, like, as up to date on things as Greg, so for me, it's kind of fun to just be looking at these things for the first time and being like, oh, okay.
That sounds really interesting. And, I mean, I personally enjoyed it a lot.
I have no idea what You can be honest. You can tell me, like, hey. I this is this is the worst hour I've had this week.
I The best thing that ever happened to me you know what I would actually have loved? I would love to also react to video clips.
There was a there's there's some clips that I was watching today on Twitter, x, and I was like, I I would like to rip I would like to react to clips as well because there's just so so much good stuff out there.
Alright. Yeah. I agree.
I mean, I'll bring some clips next time and see what you say.
See what you say. Think we should have something to really sip. Right.
Oh, I have it's
funny you were mentioning LAN parties. So this you know what this is?
This was a ball. It's called balls. Oh, no one.
And it it was a popular drink in the LAN party era of the mid two thousands.
So if you went to a gaming LAN party in North America between North two thousand and three to 2008, you were drinking balls.
Okay. In Ireland, we were drinking black tea and beer, but fair enough.
What what so, were you a Halo man or Counter Strike or what was your what were you playing in LAN parties or what were you I was Halo Halo one, Halo two, GoldenEye, which was before that.
Those those were sort of my land party things. Although GoldenEye, I think, was just on the n n 64. I was a Counter Strike guy.
Counter Strike. Okay. Yeah.
I knew it. Yeah. I like I knew it.
Like the idea of, like, the strategies around Counter Strike and the first person shorter shooter aspect of it.
And it was also, like, the number one first person shooter at the time. Um, so I I liked it. And It was all Halo, man.
It was all Halo for me and my friends. So you used to play on Xbox?
Yeah. So we I convinced all of my friends to get Xboxes. So we're all at my parents' house.
It was, I think, 16 player LAN 16 players was the max for Halo two. Halo one and Halo two with LAN parties. And so we had four Xboxes with bedsheets in between the screens so you couldn't see the other person's screen.
It was amazing. Imagine the smell of that room. That was bad.
I can smell it from here. Yeah. Take me take me back to the smell of that room.
If anyone knows the reference. In
the chat, says, it would be very appreciated if you can reply to our comments. We look we look up to you. You have experience.
We are not big, but hope next time you reply to us. I tried to reply to some of the comments,
but I happily reply. I mean, I just opened the YouTube comments now. Yeah.
I'm having, like, I mean, next time or whenever you want, I'm, like, I'm easy. I don't have anything to do. I can sit here all day and chat, so You know what we should do?
We should turn it into, like, one of those longer shows. Like, I was chatting with your producer earlier and I was talking about some of the shows that I watch.
Sometimes they're like three and a half hours long.
We should do that. I'm up for that. I'm down.
I'm down. I don't need to pee or anything. I'll just sit here and hold it in.
Just hold it in.
The longer this is a scientific fact to everyone listening. The longer you hold the pee in, the more it just goes into your blood, and then you don't need to pee. Science.
Don't Google it. Then you do have blood, uh, pee instead of blood, but that's fine.
Yeah. He's not a doctor. He's a shrimp millionaire.
Okay, everyone? Did anyone ask in the comments what the shrimp millionaire No. It says it on my actual tie it says it on my thing.
Okay. Nice. Yeah.
No one said anything. They just assume that you're just a shrimp millionaire.
A shrimp millionaire. I guess people think I run, like, shrimp ships or something. And then they're like, why did you get this guy on to talk about AI stuff?
Okay. I'm gonna answer a few questions
from the chat.
Let's do it. What
business opportunities can one do for seniors in Canada for a person just starting out in the tech space?
Well, one thing you can do is, like, connecting young people to seniors.
Like, a dating kind of situation, you mean? Not a dating thing. More
more of, like, if you're for example, in a senior home, senior living home, like, the reality is you don't wanna hang out with seniors all day.
Like Speak for yourself. Yeah. I'm speaking for myself.
I'm speaking for myself. I think that, you know, these people want younger people to hang out with. They want events and stuff like that.
I think there's opportunities there. I think these people also wanna learn about AI too. Like, they wanna understand what ChatGPT is.
Like, I bet you can go to a senior living facility and say, hey. Pay me $50 a person, you know, and I'll teach them for thirty minutes how to use ChatGPT or something like that.
Um, there's also, like, AI optimization and transformation for senior living facilities. So I think there's a lot there.
Those are the three that come to mind. Anything that comes to your mind,
Jason? I'll be a 100% honest. When when the question first came in and when you were talking, all I could think of, I couldn't think of anything and I was like, I'm just gonna skip this, but actually what you said, so one of the events that I ran this year, the age range of the event was almost nobody was below 50, like almost nobody.
I know senior is not 50, but that's just to not the age range I thought would be purchasing things that I'm selling, and I realized that a lot of people, and what some people said to me in this age range, 50, was like, yeah, people don't really make stuff, people don't make cool stuff for us.
It's all like we go on cruises, you know, we do, there's like very specific things that people in our age range are supposed to do, and actually, I think it's really a nice and heartwarming thing that you said, this idea of actually, maybe people in that age range and for example, like, my mom is like 68, and she's using these tools and asking me questions about them all the time and really interested in them, and I think it is a shame that a lot of this stuff is just targeted at a very, very young audience, considering the entire world is getting much, much older now.
So actually, I think that's a very lovely idea where you look at what are the things that are being taught, like, by mainly to younger people, and would there be a way to make that into an experience that people, you know, people who are much older would actually enjoy, and it's also customized to them.
So for example, they might not want to do a virtual version of it. They might not want to sit down for six hours to learn something. Maybe there's a way to do it that's more interesting.
Actually, I really think that's actually a very nice thought, but maybe the answer to the question is why don't you just not consider them to be a different type of person and also advertise to them?
So I spent some time actually working on products for older adults, and one of the biggest insights I had was that older adults don't see themselves as older adults. So people who create products for older adults, in the marketing material, they have, like, actors and actresses that are have white hair and who have tons of wrinkles and who don't look really good and stuff like that.
They don't see themselves like that. They they you know? So, you know, what we what you can do is you could and set let's say you're targeting 65, like, have 45 year olds and 50 year olds and 55 year olds there.
Make have them, you know, smiling. Like, these are it's sort of unfortunate that the smartest people in tech build products for, like, Gen Z and millennials.
Yeah. Because it I just believe that one of the greatest underserved markets is products for older adults that make them happier and healthier.
And I think that the missing link that a lot of people have when they're actually creating these products is don't position them as, like, old people products.
Yeah. Please don't. Don't.
Okay.
I'm gonna go into chat. Greg, sorry. Sorry.
Can I just based on what you said there, I think a lot of people listening to your podcast, they are entrepreneurs, they want to start businesses, they want to grow their businesses? I think one of the things that Silicon Valley does, which is not great, is that it makes you think that the only type of business that you can build is a crazy scale up that has to make millions.
If you were able to build something that just makes 5 or 6 k per month and it covers all of your expenses plus, you know, some profit for whatever you need, looking at whatever salary you need, maybe you need to make 120 ks a year, then you can actually start considering business models that are just a lot easier and more enjoyable that don't have to be super complicated, and if you think about like, okay, I'm just going to focus, if you're thinking about seniors in Canada, don't even start there, I'm going to think about the seniors in my city and what I can do here, and if I feel like I want to have more, then I'll expand more, but it doesn't always have to be this massive vision, mission and values, huge statement.
I can tell you most of my businesses are just they have no vision statement, no mission, nothing. It's just a small thing that makes some money, and I don't need it to be anything else, and I think the senior thing and the thing you said, you know, they don't see themselves as older people, but being able to do something for a small group of people first, not making it into some crazy thing, it's just so much more relaxing to start a business like this.
More comments, what's good, Ank? We've got Are we Ank's?
We're Ank's. We have someone saying, I love shrimp, so we got that.
Shrimp. Greg, do you review creator tools? We created tool for hooks.
Could be cool in the future to review actual tools on this livestream. Yeah. And that's about it.
I think we can end this dream and make it short and sweet. What do you think?
I think we should just awkwardly stay here all night and, uh, never never stop.
No, I'm good. If this has been remotely interesting, please like, comment, reply, repost, do something so we know to do this again.
I don't know if this is fun for other people and useful for other people. I hope it is. This is we're we're learning as we go in real time, and we'll be sharing sharing what we learn in real time.
Thank you all for spending time with us. This is an incredibly exciting time to be on the timeline, to being agree to agreeing with things, to disagreeing with things, to trying things, to getting your hands dirty.
So I have the utmost respect for anyone who's out there building, doing, having fun, and, you know, I'll see you I'll see you on the next one.
Thank you, Jonathan, J Ice Cream, for joining me as my cohost today, and I'll sip to you.
Add Modern Creator as a preferred source and Google shows you more of our breakdowns in Search, Top Stories, and AI Overviews. It only changes what you see, and you can undo it in your Google settings anytime.
Add to Preferred SourcesOpens your Google source preferences with us pre-loaded. Tick the box and you're done.
A solo walkthrough of five open-source GitHub repos getting traction right now, an AI writing editor, an agent-run CRM, a video-editing agent, a skill security scanner, and a phone-controlling harness, each with the exact install command and the first small workflow to try.
Greg Isenberg names the role he thinks AI agents are about to make the most valuable job in tech, and hands over the folder structure, tool stack, and 30-day plan to become one.
Greg Isenberg and developer Vinny break down WebMCP, the experimental browser feature that lets any AI agent search, compare, and buy on a website without scraping the page, then price two startup ideas built on it.
Cloudflare quietly built the payment rail for AI agents to pay tiny fees for web resources, and that rail is about to spawn a whole generation of businesses selling clean data, agent-readiness, and packaged expertise.
Greg Isenberg's nine-part framework for turning Claude Code into a 24/7 AI employee, walked through live on a real med-spa idea, from workspace and memory files to scheduled routines and parallel agent sessions.
01:35
02:39
03:42
04:46
05:50
06:53
07:57
09:02
10:04
13:15
14:19
15:23
16:26
18:34
19:37
20:41
21:45
22:48
23:52
25:59
27:03
29:10
30:14
31:18
32:21
33:25
34:29
35:32
36:36
37:40
38:43
39:47
40:51
41:54
42:39
44:01
45:05
47:12
48:16
50:29
52:31
54:38
55:42
56:45
57:49
58:53
59:56
61:31
61:53
64:31
65:15
66:18
67:22
69:29
70:33
71:37
72:40
73:44
74:48
75:51
76:55
77:59
80:06
81:10
82:13
83:17
24:43
35:18
28:58
44:21
34:10
48:10