📺 Today’s recommended deep-dive video: https://www.youtube.com/watch?v=6qLoYDZJpdk
The Shrimp Millionaire’s Guide to AI Arbitrage and “Anti-Shittification”
Greg Eisenberg and “shrimp millionaire” Jonathan Courtney deconstruct the most controversial business takes of the week, moving beyond news to uncover high-margin opportunities. From million-dollar baby food apps to the death of the “two-pizza” rule, they explore how humans remain the ultimate differentiator in an automated world.
Core Question: How can entrepreneurs leverage AI to build lean, high-margin businesses while maintaining the human connection that customers are increasingly willing to pay for?
Highlights
- The Trust-Price Paradox: Why parents (and pet owners) often trust $100 apps more than free alternatives when safety is at stake.
- The Human Premium: How “anti-shittification”—doubling down on human support—has become a massive marketing advantage for modern brands.
- The Death of the Two-Pizza Rule: Why the ideal startup team size has shrunk from twelve people to fewer than seven in the age of AI.
- System Psychosis: The danger of CEOs spending $5,000-per-hour time building $5-per-hour AI automation loops instead of selling.
⏱️ Reading time: approx. 8 minutes · Saves you about 77 minutes vs. watching.
Want to take notes while watching? Click the image below and let AI Notebook capture the key points for you 👇
High-Ticket Niches and the Baby Food Goldmine
The Psychology of Parental Trust
The conversation kicks off with a startling statistic: a baby food app, Solid Starts, is generating $1 million in monthly revenue. For many parents, the first 100 days of a child’s life are a period of intense anxiety, particularly regarding health and nutrition.
Jonathan Courtney points out that in high-stakes markets like weddings or infant safety, price acts as a proxy for quality and commitment. Parents are often more likely to trust a high-priced application because they perceive the cost as a guarantee that the company is taking safety seriously, rather than just delivering AI-generated fluff.
This creates a massive opportunity for entrepreneurs to specialize. If a general baby food app earns seven figures, a niche version for specific geographic regions or even “Solid Starts for Puppies” represents a viable, lower-stress business model.
💡 Digging Deeper
Q: Why do these apps work better when they are more expensive?
A: In “one-shot” scenarios like parenting or weddings, customers use price to filter out low-quality noise; a high price suggests a professional team is auditing the health data.
Q: How can this be applied to other niches?
A: Any market where “safety” or “high stakes” is the primary concern is ripe for this; think pet health, elderly care, or specialized medical diets.
Q: Is there still room for newcomers?
A: Yes, specifically through “geographic nicheing” or hyper-specialization, such as apps for babies with specific allergies or regional dietary habits.
The Rise of “Anti-Shittification” and Human Support
Reclaiming the Customer Experience
There is a growing trend known as “inshittification,” where dominant companies slowly degrade their product quality and support to maximize margins. Greg and Jonathan argue that the “next big thing” in marketing is explicitly rejecting this trend by guaranteeing human-only customer support.
While engineering teams are being “eaten” by AI—meaning customer support reps can now use AI to fix bugs or prototype features—the human element remains the premium differentiator. Companies like Basecamp and Trade Republic are winning by putting real faces and real voices at the forefront of their service models.
Startups that automate 100% of their support are often falling into a trap. Customers, especially those buying high-end products like the Eight Sleep mattress, expect a human to intervene when things go wrong, making “Human Support” a powerful luxury brand pillar.
💡 Digging Deeper
Q: What does it mean that “support is eating engineering”?
A: It means LLMs allow non-technical support staff to analyze feedback and generate functional code prototypes, shortening the loop between customer pain and product fixes.
Q: Can a company really scale with 100% human support?
A: It’s a choice of margins; the “human premium” allows you to charge more, which offsets the higher cost of staffing compared to a bot.
Q: Is AI support always a bad idea?
A: Not for low-tier troubleshooting, but it becomes a brand liability the moment a customer feels stuck in an infinite loop with a machine.
Small Teams and the Death of the Two-Pizza Rule
The New Startup Physics
For twenty years, Jeff Bezos’s “two-pizza rule” suggested that teams should be small enough to be fed by two large pizzas—roughly 12 people. In the era of AI, David Pan (CTO of Cursor) suggests that two pizzas are now far too many.
The modern startup doesn’t need a dozen people to reach $10M in revenue. With AI agents handling SEO, code generation, and initial marketing outreach, the ideal team size has collapsed to somewhere between one and six people.
Jonathan argues that before hiring your “savior” employee, you must first define their tasks and attempt to build an AI agent loop to handle them. Hiring is no longer the first solution to a scaling problem; it is the last resort after automation has been exhausted.
💡 Digging Deeper
Q: Is a one-person startup actually realistic?
A: It is for a technical founder, but usually, a “pizza for one” needs a partner—one person to build the product and another to promote and sell it.
Q: What is the biggest mistake founders make with AI agents?
A: Trying to build one “god agent” that does everything; the transcript suggests using specialized sub-agents with limited “radio of blast” for specific roles.
Q: How does this change venture capital?
A: It makes high-margin, “asset-light” businesses more attractive than traditional high-burn startups that need 50+ employees to function.
Fighting “System Psychosis” and the Loneliness Epidemic
The CEO as a Promoter
Many founders are currently suffering from “system psychosis,” spending months building intricate AI systems to summarize meetings instead of talking to customers. Jonathan warns that a CEO’s primary job is to make people aware that their product exists.
No amount of “perfect” internal systems can save a company that fails to market itself. Marketing is often viewed as “dirty” in Silicon Valley, yet the most successful founders—from Elon Musk to Sam Altman—are essentially full-time promoters who spend their days on podcasts and social media.
This extends to the physical world, where “third spaces” like bowling alleys and bars are disappearing. There is a massive arbitrage opportunity in using digital audiences to fund and fill premium physical locations, such as luxury arcades or community-focused “Vogel-style” cafes.
💡 Digging Deeper
Q: What is a “third space” opportunity?
A: Since traditional social hangouts are closing, building a premium, “nostalgic” physical space (like a high-end LAN party venue) for a pre-existing digital audience is a massive opening.
Q: How do you avoid over-engineering your business?
A: Follow the “revenue first” rule; don’t build a system to automate a problem until the problem is actually causing the business to lose money.
Q: Is “Marketing” just another word for “Gimmicks”?
A: No. Gimmicks are expensive one-offs; marketing is the consistent work of making your product known through storytelling and presence.
Key Takeaways
The current business landscape is defined by a paradoxical shift: as AI becomes more capable, the value of human-centric services and small, agile teams skyrockets. Entrepreneurs who focus on high-stakes niches—where trust is more important than price—can build incredibly profitable businesses with fewer than half a dozen employees.
The greatest danger today is not being replaced by AI, but becoming obsessed with the mechanics of AI at the expense of sales and human connection. Whether you are building a baby food app or a luxury bowling alley, the winning formula remains the same: use digital tools for efficiency, but keep a human at the center of the experience.
Finally, remember that “anti-shittification” is a viable business strategy. In a world of automated bots and declining service quality, being the company that answers the phone with a real person is no longer just good service—it is a premium marketing asset that justifies higher margins and builds lifelong customer loyalty.
Q&A
Q1: What is the “shrimp millionaire” joke about?
Jonathan Courtney jokingly claims to have made his millions in the shrimp business, though he is actually the founder of AJ&Smart, a prominent design sprint agency. It’s a running gag about how anyone can look like a millionaire in any niche.
Q2: Should I use AI for my customer support?
The consensus is “Sip” on using AI to help your human agents work faster, but “Skip” on replacing humans entirely. High-end customers perceive 100% AI support as a sign of a cheap, “inshittified” brand.
Q3: Is the “Third Space” business model too expensive for tech founders?
While physical businesses require more capital, the trick is to use “digital-first” funding. Build an audience on X or Instagram first, then sell memberships or pre-sales to fund the physical location, reducing the risk of a bank loan.
Q4: How do I know if I’m suffering from “System Psychosis”?
If you have spent more time in the last month building AI automations for your internal workflow than you have spent talking to potential customers or closing deals, you are likely over-engineering.
Q5: What is the “arbitrage” opportunity in AI right now?
It’s the gap between those who know how to use AI for specific tasks (like SEO or content) and traditional industries (like a local tire shop) that haven’t adopted these tools yet. Being the first in a “boring” niche to use AI gives you a massive, temporary advantage.
Q6: What is a “sub-agent” in the context of an organization?
Instead of one AI trying to run your whole marketing department, you create a “Thumbnail Agent,” an “SEO Research Agent,” and a “Transcript Agent.” Each has a narrow role, making them more reliable and easier to fix when they break.
Q7: Is it true that parents trust more expensive apps?
In many cases, yes. When it comes to infant health, a “free” app can feel untrustworthy or data-hungry. A premium price tag suggests the data is curated by experts and the company is incentivized to keep the child safe.
