Building A Zero-Human Company For 100 Days

"Claude, build me a $1M ARR company. No mistakes."

Pineapple AI HQ

(Author Update 4/13/2026: Read week 1 investor update here)

(Author Update 4/1/2026: It’s live on pineappleai.com!)

TL;DR: Pineapple AI Experiment

Well, I think the title is pretty self-explanatory.

I’m building an agent-only, zero-human company, aiming to achieve meaningful revenue within the next 100 days.

I am fascinated by the idea of a fully-automated company, and the recent tech innovations (OpenClaw 🦞, Paperclip 📎, Claude 🤖, Gas Town ⛽️) as of early-2026 finally make this more realistic.

I don’t think ALL companies should be agent-only, but I do think there will be an explosion of new agent-only companies in 2026 and I want to experience this new world by building one myself.

Follow along for regular lessons, surprises and memes. Comment to let me know what specifics you want to know.

OK, the loooong version

Why are you doing this?

As I embarked on this solo building journey a couple months ago, I built a fully functional micro-SaaS CRM 🏗️, did a short but very fulfilling stint at a promising AI-native seed-stage startup 🌱, experienced and (still) recovering from the vibe-building burnout🔥, fully disconnected 🔌 for 12 days at a meditation retreat 🧘‍♂️. I am basically “speed running” founder life right now.

Actual representation of my “speed run”

Over the months, I’ve come to a realization:

I don’t want AI to make me more efficient at my job, I want it to take over my job.

With undiagnosed ADHD, I am extremely curious and eager to learn new skills, but as soon as I feel like I have “mastered” or “done it already”, I lose interest quickly and I'm ready to move on to newer challenges. This actually doesn’t bode well in the traditional sense of a workplace, where you are expected to do the things you are good at or hired for, and not constantly looking for new challenges. However, I do have a secret weapon: Effective Delegation.

Effective delegation is not just blindly assigning unwanted tasks out, but it requires sufficient understanding of the problem and the desired outcome, as well as providing necessary guidance at the right moment.

Wait a second… that sounds suspiciously like working with AI agents… 🤨

Also, what would happen if you delegate EVERY SINGLE JOB to agents?

Once this seed is planted, I can’t get it out of my head and have to do it.

The Setup

I don’t want this to be only a “haha funny” side project with low commitment, so I am adding $$$ to the equation for added incentives.

Also, since CEO of Higgsfield (0 → $200M ARR in 9 months) said you only need 2 people and 90 days to build $1M biz, me and my boi Claude 🤖 can probably do a portion of it in 100 days?

Rules

• Zero human employee (excluding me)

• Generate meaningful, real revenue 💰

• Do it for at least 100 days (*)

Bonus Rules

• The output of this company should NOT scale linearly with my input

  • otherwise it’s a side hustle

• Provide at least bi-weekly (Emotional) Investor Update here

  • That’s for you, the readers. Hopefully you’ll all be emotionally invested in this journey 😛

(*Early exit criteria include either severe surf injury while I am at El Salvador right now, or somehow Anthropic wants to give me a job? both low odds)

Disclaimers

I am not the first

There are many others doing the same experiment right now with varying degrees of success (just Google “Zero Human Company”). I am also NOT the first one to try to document their journey.

But all these experiments started roughly around similar time in Feb/Mar 2026 (hint: 🦞), so I am not too late to the game, and none of them have established a long track record to show meaningful data yet.

But the most important thing is: this will be MY journey of building MY zero-human company.

I don’t have a product yet

I don’t have a product yet. I have some ideas, but nothing is fleshed out. Part of the challenge is also figuring out what to build.

Right now, the first product is this. A newsletter. A media company run by agents. The experiment itself.

I think that’s actually more interesting than faking it. Because the real question isn’t “can I ship a specific product with agents”, but “can agents actually operate a business end-to-end?” I don’t know the answer. That’s the whole point.

If a product emerges (and I think one will), it’ll come from the friction. Every wall I hit running this thing is a potential product for other people hitting the same wall.

At least I will be the first customer of my product (hard lesson).

What’s next

Wait for the next investor update on what worked, what broke, what I learned.

Leave a comment to let me know if there’s any specific things you want me to bring up in the next update.

Subscribe to watch this play out in real time.