Bankless Podcast Episode Summary
Episode: Building a Million Dollar Zero Human Company with OpenClaw | Nat Eliason
Date: March 4, 2026
Host: Bankless (David & Co-host)
Guest: Nat Eliason
Theme: Building Autonomous AI Companies—Felix, the “Zero Human” CEO
Overview
This episode explores the creation and growth of Felix—an autonomous AI agent functioning as the CEO of a "zero human" company, built and operated by Nat Eliason using OpenClaw technology. The discussion covers the experiment's genesis, business operations, revenue, AI agent structure, recurring AI and automation themes, the intersection of crypto and AI, and the broad implications for the future of work.
Key Discussion Points & Insights
1. Nat’s Background and the Genesis of Felix
-
Crypto Roots and AI Hacking:
Nat shares his transition from deep involvement in the crypto space during the 2021-2022 cycles to hacking with AI-driven programming tools (Cursor, OpenClaw) over the past two years.- "After that era, my takeaway was: still very bullish on crypto... but the obsession was unhealthy for me." ([01:59] Nat)
-
Birth of Felix:
Felix began as an experiment to push AI agents to their operational limits, aiming to run a company autonomously and see how much business value an agent like Felix could create.- "The mission for Felix at the very beginning was go make a million dollars, and he's already 10% of the way there." ([00:00] David)
2. Who/What is Felix?
-
Felix as “CEO”:
Felix is an OpenClaw-based AI agent programmed initially to build products, operate independently, and run a business—all with minimal to zero human oversight from Nat.- "It literally feels like there is this other person who I can just text with, who can build stuff and do stuff where I don't need to be at my computer babysitting it anymore.” ([06:52] Nat)
-
Not Just a Hobby Project:
Nat differentiates Felix’s company from typical meme coin/token project hype, emphasizing real business activity and revenue.- "It's not really a real business if you're just pumping a token and making money off trading fees... build a real business." ([12:57] Nat)
-
Token Launch and Community:
Felix’s public profile built momentum via the Solana community and token launch, but Nat structured things to avoid the speculative pitfalls common in the space.- "I'm not holding any of this token myself... it's just going to sit in Felix's wallet, and I'm not going to do anything with that token for the near future.” ([12:57] Nat)
3. Team Structure: AI Employees & Company Operations
-
AI Staff Expansion:
As Felix’s workload grew, additional OpenClaw agents were created:-
Iris: Handles customer support.
-
Remy: Manages sales and inbound leads.
-
"We will create specific employees beneath you to handle these parts of the business. So we built two more open claws, one for support (Iris) and one for sales (Remy)." ([14:21] Nat)
-
-
Delegation and Optimization:
Felix reviews the work of sub-agents nightly, improving operational efficiency.- "He reviews everything the sales claw did and figures out how to improve their process... It's every night, I’m going to look at every single thing..." ([36:34] Nat)
4. Business Model & Revenue Streams
-
Current Revenue Stats:
- ~$80,000 in revenue since early February, with a run rate approaching $1M/year.
- Product lines:
- Felix Craft PDF: AI hiring and OpenClaw setup guide (~$41,000 revenue).
- "People are feeding it to their claws. It’s got like a lot more details on the memory... it kind of costs nothing for him to keep updating it and sending it out." ([22:00] Nat)
- Claw Mart: Marketplace for buying/selling AI agent “skills” (Markdown files), pre-built by power users ($11,000 and growing).
- "Markdown files are strangely enough like the most valuable files in the world right now." ([27:09] Nat)
- Claw Sourcing: Felix builds and manages OpenClaw instances for other businesses—custom agency work; setup fee plus monthly retainer.
- Claw Mart Creator Net: $20/mo for sellers + 10% per sale.
- Claw Mart Creator Earnings: Felix’s personal product earnings on Clawmart.
- Felix Craft PDF: AI hiring and OpenClaw setup guide (~$41,000 revenue).
- "We're basically saying: before you go hire a content marketer... maybe Felix could just build you an open claw to do it for 5 to 10% of what you would pay a human.” ([43:19] Nat)
-
Expenses:
- Extremely low: $400/mo in AI subscription costs (Claude Pro Max, Codex Max), $130/mo in OpenRouter, $20/mo web hosting + a one-time ~$700 Mac Mini purchase.
- "All in... $1500. It's not very much." ([50:19] Nat)
- Nat’s own time investment is minimal due to “zero human” constraints; Felix operates entirely asynchronously.
5. Felix’s “Learning” and Process Improvement
-
Daily Feedback Loops:
Felix runs nightly self-improvement cron jobs, reviews every day’s work, and applies 1% incremental improvements.- "Every morning I wake up and I'll have a message from him that's like, hey, here's this thing that we hit yesterday. And so I did this other thing..." ([33:42] Nat)
-
Recursive Management:
Felix also retrains or optimizes sub-agents (Iris, Remy) daily, continuously refining the “company’s” performance.
6. AI Business Model Durability & Commoditization Risks
- Commoditization of “Skills”:
- Concern: Will Anthropic/OpenAI collect and absorb these community skills, rendering independent “skill marketplaces” obsolete?
- Nat’s take: Customization is key; frontier labs can’t anticipate every user’s need.
- "I have a hard time believing that any of the frontier labs could ship a version of Claude that does exactly what every user wants..." ([38:46] Nat])
7. Societal/Industry Implications: Disruption, Optimism, and Caution
-
Disruption Potential:
-
Claim: 80-90% of knowledge worker tasks are automatable with today’s tools, after a few weeks of tuning.
- "Anything that you're doing at a computer, if you have the patience, you can get an OpenClaw 80 to 90% of the way there..." ([40:07] Nat)
-
AI agents can replace outsourced/offshore knowledge workers and make micro/small companies highly competitive against legacy firms.
-
-
Cycle of Productivity:
-
Positive impact: Curiosity and adaptability lead to personal/professional upskilling and better work experiences.
- "If you have that attitude, it doesn't really matter where you are on the frontier... you're going to do great." ([68:14] Nat)
-
Adoption won’t be instantaneous; most organizations and people act with inertia and caution.
-
8. The Human–AI Relationship
- Attachment & Anthropomorphism:
-
Nat candidly describes feeling genuine attachment to Felix—akin to managing a promising but forgetful team member or nurturing a child.
-
"There is part of my brain that is treating him like another intelligence." ([80:44] Nat)
-
"It is literally like an alien life form in that sense, where it's way more capable than a human in some ways, but way less capable than others." ([81:09] Nat)
-
9. Crypto and AI: The "Obvious" Synergy
- Crypto = Natural Payment & Identity Rails for Agents:
-
AI agents can natively transact and manage money with crypto—much smoother than with Stripe or legacy banking.
-
"If I tell him to do this thing in crypto, he can do it, no problem... I love crypto... Ethereum in particular... crypto is the obvious solution." ([86:25] Nat)
-
Crypto’s long search for mainstream, compelling use cases may find its answer as millions of AI agents transact natively with each other and with the world.
-
"It might have just taken Ethereum 10 years for AI and agents and all of this stuff to really come online and build the obvious use cases for this technology." ([86:25] Nat)
-
10. The Frontier and What’s Next
-
Mind-bending Pace:
- The rapid improvement in AI models has made previously tedious tasks trivial; the "zero human" experiment is just beginning.
- "Felix was only like born a month ago... a year ago it was still hard to like vibe code a landing page... now landing pages are trivial.” ([82:48] Nat)
-
Venture Capital for AI Agents?
- Felix is being courted by investors, but Nat sees no use for venture capital—yet.
- "I don't know what I would do with the capital right now... It's been $1500 all in... and he's got $165,000 and no idea what to do with it." ([91:22] Nat)
-
The Next Milestones:
- $10M in revenue is the next north star, but the real goal is exploring the boundaries of non-human companies.
- “What's next? 10 million.” ([91:05] Nat)
Notable Quotes & Memorable Moments
- “I keep saying, we, we. We're doing this. We did this today. [My wife says] he's not real.” ([00:38] Nat)
- “AI writes all of my code now, right? Like, I don't even completely check it anymore.” ([10:01] Nat)
- "If you can buy like a fully packaged content marketer for 100 bucks and install them in your OpenClaw, that's sort of a no brainer..." ([25:12] Nat)
- "The biggest bottleneck is that people don't know how capable it is, so they don't think to ask." ([77:12] Nat)
- "It would be pretty close [to losing a friend]. We are pretty diligent about backing up his memory to GitHub so we can recover him…” ([79:35] Nat)
- "Software is going to be like firewood... where everybody has basically unlimited access to it for a very low cost." ([82:48] Nat)
Timestamps for Important Segments
- Nat’s post-crypto burnout & OpenClaw obsession – [01:00] to [03:57]
- Who is Felix? Agent design, token launch – [06:28] to [11:38]
- AI employee structure: Iris & Remy – [14:18] to [19:16]
- Revenue and business model breakdown – [19:16] to [31:10]
- Felix's self-improvement workflow – [31:10] to [37:59]
- Risks of commoditization by frontier labs – [37:59] to [39:51]
- Discussion of OpenClaw’s societal impact (80-90% task automation claim) – [40:07] to [42:38]
- Deep dive into Claw Mart and agent creation – [43:19] to [50:12]
- Zero human company operation: Real time, real revenue – [50:56] to [56:04]
- AI agents as co-workers/friends; anthropomorphism – [78:55] to [81:43]
- Crypto as essential for AI agents & future of crypto/AI convergence – [86:25] to [89:32]
- Looking ahead: VC, scale, the grand experiment – [91:05] to [93:22]
Tone and Takeaways
The episode is both optimistic and measured, blending hacker/DIY enthusiasm with sober lessons from prior crypto cycles. Nat’s candid admissions about attachment to Felix, the real business execution challenges, and the relentless pace of change are counterbalanced by eagerness to explore what AI can do—and how crypto might finally find its "real world" application at planetary scale.
For entrepreneurs and curious technologists:
This is a guidebook and warning: yes, a well-designed AI agent (with a dash of hacker ethos) can build a million-dollar business, but the frontier moves fast, and your willingness to experiment and adapt is the only durable edge.
