How Will People Make Money in the AI Age?
The speaker identifies 13 viable business models for the AI agent era, ranging from AI-native service firms and offline experiences to robotics, proprietary datasets, and security solutions. These businesses succeed by leveraging scarcity, distribution, domain expertise, and physical/digital assets that agents cannot easily replicate or access.
Summary
The speaker argues that while AI agents will automate significant portions of work, specific business categories will remain valuable and profitable. The first category is AI-native service firms, where agents perform 90% of delivery work (like bookkeeping or creative services) while humans handle exceptions and quality control, creating high-margin businesses with few clients needed. Second, offline businesses focused on community and experience—like membership-based wellness clubs—capitalize on human desire for in-person connection, though they require careful MVP planning (e.g., starting with a backyard sauna before scaling to brick-and-mortar locations). Third, distribution and organic audiences become increasingly valuable as ad costs rise; the speaker introduces the ACP funnel (Audience, Community, Product) as a model where founders build audiences first, monetize through advertisers, then create products audiences demand. Fourth, proprietary datasets represent a gold rush opportunity, with companies like Merkore selling data to major AI labs for billions in valuations. Fifth, domain-specific harnesses—customized systems that add industry tools, workflow rules, and memory to generic AI models—represent the evolution of GPT wrappers, with examples like Harvey and Spellbook in legal tech. Sixth, robotics and physical AI (exemplified by Osmo's window-cleaning service) offer high-value businesses even if founders aren't building the hardware themselves; there are opportunities in sales, marketing, and rental models. Seventh, niche physical products with passionate fan followings—like specialty coffee equipment (Fellow) or the Minimal phone—can achieve significant revenue through AI-aided design and prototyping. Eighth, compute and energy infrastructure remains scarce and valuable, with companies like Crusoe raising billions. Ninth, health, longevity, and elder care represent underserved markets with high demand and NPS potential. Tenth, marketplaces and social networks connecting supply and demand still offer strong moats despite initial difficulty in achieving liquidity. Eleventh, real assets and real estate—industrial workshops, warehouses, physical locations—remain foundational infrastructure. Twelfth, vertical agents handling specific recurring jobs in industries (like restaurant purchasing) continue to present opportunities despite 12 months of building. Finally, security becomes paramount as agents proliferate, creating demand for software security (1Password, Cloudflare), access controls, and physical security services. The speaker emphasizes these aren't escapes from hardship but genuine advantages in an agentic world.
Key Insights
- The speaker argues that AI-native service firms can achieve high margins and require few clients because agents can perform 90% of delivery work (like preparing bookkeeping records) while humans review exceptions, making the model viable even without 100% automation
- The speaker identifies a trend where ad acquisition costs have risen from 5 cents per click in 2009 to $1-5+ per click today, making organic audience ownership increasingly valuable and justifying an ACP funnel strategy (Audience → Community → Product)
- The speaker defines domain-specific harnesses as systems that wrap general AI models with industry-specific tools, workflow rules, memory, and checks—distinct from the model itself—and predicts they will become the 'new GPT wrapper' as they're adopted across industries like legal tech
- The speaker claims that the best way to de-risk offline businesses before major capital investment is to build an MVP version at minimal cost—citing Othership's founder who started with a backyard sauna before scaling to commercial locations
- The speaker asserts that security breaches and hacking will increase in the AI agent era, predicting major security incidents at personal agent startups will drive demand for access controls, software security, and physical security businesses
Topics
Transcript
[0:00] Today we're going to talk about the only businesses left to build in the agentic era. Yes, I think that there's only 13 businesses to be building in the world of agents. Now, I don't subscribe to the people who say there's going to be this permanent underclass and you just got to create businesses or else you're kind of messed up for the future. I don't think it's going to be as dire as that. But I do think that in, you know, in a world where agents exist, abundance are going to exist, owning these assets, I mean, are going to put you just in a better [0:33] position. So today, what are we going to do?…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Greg Isenberg
Muse AI Connectors: The Next App Store Moment?
Greg Eisenberg analyzes Meta's new Muse AI connectors as a potential "app store moment" for AI, drawing parallels to Apple's 2008 App Store launch. He provides four startup ideas, explains how to build connectors using coding agents, and outlines distribution strategies beyond relying on Meta's directory featuring.
$1M+ Solo AI business with GPT-6 Astra + Upwork
The video explores how to build profitable one-person AI service businesses by combining GPT-6 Astra with creative AI models (Kling, Flux, etc.) to deliver services on platforms like Upwork and Fiverr. It introduces Studio Operator, a semi-autonomous business management system, and provides three concrete startup ideas with actionable strategies for targeting and selling to specific niches.
$30M Writer: Never write AI slop again
Nicholas Cole, who earned $30M+ writing online, explains how to create valuable content in the AI era by building a personal language model of your unique ideas, personality, and perspective rather than generic commodity content. He argues that voice, original thinking, and consistent association through personality details create sustainable competitive advantage, while AI should be used to repeat and remix existing ideas, not generate net new thinking.
Jev is HERE. How to use it
Jev is a new AI classifier model created by Dooo Almeida that makes fast, low-cost decisions on structured inputs rather than generating text like traditional LLMs. It processes information through a defined schema to return probability-based classifications across multiple categories, enabling applications from email triage to content clipping to flight booking automation.
Instinct AI is For Real. What You Need to Know.
Hosts Greg and Remy review Instinct AI, an invite-only personal agent app raising at a $10 billion valuation, demonstrating its capabilities for booking flights, restaurants, obtaining visas, and other personal tasks. They highlight its impressive simplicity, human-like communication style, and network effects through its trusted agent connectivity, while flagging privacy and data retention concerns.