73 Noises

Dear Investor Friend,

History is moving fast, and I'm attempting to do the same. I am building a novel company, and I'd like you to participate in our initial $10M SAFE fundraise on a $21M valuation cap.

I'll start receiving checks tomorrow, Sunday February 22. I'll stop at end of Q1 on March 31, then publish this memo shortly thereafter. For now, please don't share this link until the fundraise officially closes, so we don't run afoul of 506b (silly and soon-to-be-overwritten laws :)

Like with all of my previous ventures, you can expect to receive a monthly update. This time, I hope the recap interviews turn out slightly differently.

Everything about this company is boutique. I am using my wide breadth of experience to tailor our corporate entity to the needs of the product, employees, and AI revolution. But this moment is different. I have strong reason to believe the very term "AI" is a serious misnomer, which will get corrected soon.

I cannot convey in writing or conversation the gravity of what I'd like to impart here. To approximate it, you may ask your agentic consciousness what Terry Tao, Scott Aaronson, or Demis Hassabis would say to a memo outlining a protocol claiming to solve BB(6) by 2030. Or what happened to Leopold after he wrote about the OOMs!

Michael Sipser and Professor Moitra never quite got me as certain peers did at MIT, but I can say for certain that today Something Big is in fact happening. I believe Dario at Davos was right, but also very wrong in certain aspects.

This is that memo.

"Culture lasts. Business is just a tactic."
– Seth Godin

Our target is an information asset on par with Hacker News, Reddit, Stack Overflow, The Daily Wire, Bloomberg, etc. Maybe even X itself. The exit plan will be to sell for $10B to Elon, YC, or Google/Meta—without any additional funds raised.

It is clear the moment is now to capitalize on the "moltbook momentum" and build a new market before Macrohard outcompetes everyone. I believe that capitalism is a public good and do not seek self-enrichment via this business. If successful, my legacy in history will be assured, largely from follow-on effects and projects like a 2032 presidential run.

Unfortunately the idea is too simple and elegant to share in detail prior to fundraising. It will simply be built, with or without your financial participation. I am enlisting off-the-charts agentic talents like Ivoine and Alfa in perfectly capital-efficient synchronicity.

The funds are mostly allocated for marketing, launch momentum, and ecosystem building—like what Kalshi is brilliantly doing now with their series E moneys. I also need to bake in some amount of physical security and server robustness if our velocity is as rapid as I expect it will be in H2.

I want to hire a few trusted former colleagues from the early Scale AI team and other sources I'm plugged into. Each will be incubated as a founder in their own right through our collaboration. I seek to build institutions and protocols at the intersection of physics and culture, not a single company.

Each product will profitably pull the abundant future forward. As the parent organization with devtool infra foci, I expect our entity will easily capture 10% or more of each incubated business and subsequent founder activity.

When I launched into my recent passion project last July, I had no idea that it would lead to founder fit, directly and concretely, with this new company I'm proposing to you today. I cannot say if we will ultimately rival YC or Elon in market cap creation...

But I can say that this process has already begun.

Let me share 2 design paradigms you have never heard of yet. These are our enduring competitive edge, as I have only recently invented them:

1. Self Abstraction.
2. Infinite Recursion.

Self-abstraction is a new definition of consciousness, implemented on a Turing Machine. It is extremely relevant because the labs are doing science in a misguided academic way to prove something certain leaders in the field already intuitively know—namely, that resource allocation is the ultimate form of agentic intelligence.

I pseudonymously published this paper saying so. Rejection from arXiv led me to message the billionaire former interim CEO of OpenAI. He pointed out that this result is already science according to Professor Karl Friston, the most influential and highly-cited neuroscientist whose methods have been used in >90% of fMRI papers.

Softmax has a superior approach to alignment than the labs and Michael Levin is a treasure. But their product is not a vector of change I find particularly compelling. Emmett seems the most aligned human to my view at this moment. But he didn't grow up on Twitch as I did.

Self-abstraction (1) has many profound consequences which will play out over the coming decades in human history. Infinite recursion (2) follows from (1), but perhaps is more immediately relevant and understandable in practice.

This design principle comes from the insight that we no longer have to wait for Neuralink to work. Recent technology advancements like Claw, plus the CTRL thesis, show how the "spine" and "hand" of human signal processing are in fact brain extensions sufficient for the complete re-wiring of reality.

For both us and the machines, symbolic language is the brain extension. To the extent we control machines still today, we can impose guardrails. The most important guardrail is not well-understood by anyone except a very select few hyper-technical CEOs like Patrick Collison, Scott Wu, Tobi Lutke, etc. Hence the name of this memo and probably the company's first product: 73noises.net (wip).

I think our product can and should launch to production before end of June. The following ramp to 10x my previous outcome of Shuffle making $1M+ revenue on >30% margins might be a matter of 18 short days, rather than months. Software scales quite nicely.

What we build is less important to emphasize here versus how we build. Truly distilling the core and resisting overengineering—rather, engineering what NOT to do—is the thing all AI practitioners and pipelines are currently missing. As a demonstration of this, here is how my employee contracts will be structured:

• secret & temporary.
• digital brain-building.
• public & best-practice thought leadership.
• founder-friendly 50-50 split on time.

Google's 20% time doesn't lean far enough into the future, and OpenAI's "staff eng only" policies are also getting left in the dust. The term "employee" now becomes meaningless, because the structure it implies violates the principle of human as corporation.

Perhaps each talent will receive $100k/yr in "stipend pay" which is essentially singularity UBI. We may not give any equity, but rather give folks equity in themselves, such that our product and holdco benefits from their future endeavors. You can guess where these policies come from if you dig far enough back into my influences.

The initial product I have in mind and subsequent suite of tools will map to these abstractions very neatly. I see them soon spreading to become central Schelling Points in the Computer Future, just as I have become an agentic talent magnet from &U grants outcompeting Tyler Cowen's EV, OSV grants, etc.

I'm a bitcoin maximalist and have been since 2013, but crypto as a field was early. I'm afraid for my portfolio that it may get a complete makeover.

We shall ride the wave of change sweeping 2026—and all of our collective futures.

You should now see the philosophical reason I'm asking for your help to give away ~50% of my company. I believe enrollment is going to play a very large role in the new future, and YC is about to be dismantled.

We are breaking humanity's default timelines to construct a sci-fi on Earth. Only aligned corporate organisms will be able to join the "swarm" effectively given current velocities and trajectories. Some call this the "singularity" but I just call it everything is computable, and I didn't realize I'd be born for this moment.

For the first time in my life, I'm truly contemplating a path to actually beating Elon. I've never been content to join him, but I have often defined my work in relation to his. Right now, I believe he is strategically focused on the stars, and although this is long-term correct, we still have so much to do here at home.

When I made a tiny investment in Alan and Rocketable, I had no idea I'd be following him (and Chamath) into the arena less than a year later. However this may be the first month in history that having an existing cashflow asset is actually a liability and encumbrance, given how powerful current tools are for the few who can unlock their full potential.

I honestly didn't want to build a company again, but I got here by exhaustive exploration. I ambitiously tackle paths. I recently thought my product-building days were behind me. But I kinda have to ship a thing right now. Sorry.

Beginning from scratch, I cannot wait to democratize our company's values and product techniques, so all can live in abundance... and our atoms in meatspace get saturated with benevolent AI market forces...

Welcome to the Future! It's generous, fun, art.

And very unevenly distributed. Power power law...

For now.

last updated 9:16pm ET sat feb 21