Skip to content
AGI Future Foundation PBC

Article 69 · Venture Studio · AGI Future Foundation

The Venture Studio, Stage by Stage: How an Idea Becomes a Company Without Losing Its Mission

How an idea becomes a company — through six gates, a mission check, and a human decision at every step that matters

Illustration of a venture studio as a journey through six garden gates along a rising stone path, from a small seedling at the first gate to a flourishing workshop city at the last gate

Illustration: the studio’s six stages as gates along a rising path. Generated illustration, not a photograph.

A new company is a bet placed in the dark. Somebody has a picture of a future that does not exist yet — a tool, a service, a way of working — and they stake months of their life and somebody else’s money on the idea that the picture is real. The venture studio exists because somebody looked at that failure rate and asked a different question. Instead of waiting for good founders to walk through the door with finished ideas, what if you built the door, the room behind it, and the team that walks through it — and what if you did that over and over, with the same tools, the same discipline, and the same refusal to fall in love with an idea before it has earned its keep?

That is the idea at the centre of this article. A venture studio is a factory that makes companies the way a good kitchen makes meals: from a standing pantry, with practised hands, following tested recipes, and throwing away the dishes that do not work before they reach the table. The studio in this programme adds two twists. First, the workers on the factory floor are largely agents — software systems that can research, draft, test, and operate around the clock — while the decisions that matter are reserved, by design, for humans. Second, the factory has a mission gate at its front door, held by a nonprofit foundation, which can stop a company that would make money in a way that betrays the mission. This article walks through the whole machine, stage by stage, from a bare idea to an exit or a wind-down.

What a venture studio is, and what it is not

The startup world already has two familiar machines for turning ideas into companies. The accelerator takes in existing teams for a fixed term, gives them a small amount of money and a large amount of advice, and sends them out to raise more. Y Combinator, founded in 2005, is the best-known example, and its public library of startup advice is one of the most-read resources in the field. The venture capital firm does the opposite: it waits for companies to exist, then buys a share of the ones it believes in. Both machines depend on founders arriving with ideas. The studio inverts the relationship. It starts with the idea, builds the team around it, and keeps the company inside the studio while it grows up.

The oldest well-known studio is Idealab, founded in Pasadena in 1996 by Bill Gross. Rocket Internet, founded in Berlin in 2007, industrialised the approach in e-commerce, and Atomic and Pioneer Square Labs are newer versions of the same pattern: a small core team, shared services, and a portfolio built one company at a time. The record of these studios is mixed, as any fair account must say. Some produced enduring companies. Some produced expensive lessons. The model is not magic. It is a discipline, and like any discipline it works only when the people running it follow it.

The discipline has three parts. First, the studio generates its own ideas instead of waiting for them, so it can aim at problems it considers important rather than problems that happen to be fashionable. Second, it shares the expensive parts — legal, design, engineering, recruiting, finance — across the whole portfolio, so each new company starts with a running start. Third, it kills ideas early and without sentiment. A studio that cannot kill its own darlings becomes a museum of half-built companies, and the stage-gate system described later in this article is the mechanism built to prevent exactly that.

What the studio is not matters just as much. It is not a fund: it does not pool outside money and invest it. It is not an accelerator: it does not take in outside teams for a term. It is not a consultancy: it does not build things for clients. And in this programme, it is not a separate entity at all. The operating model documents the studio as a function of AGI Corp — intended, not yet formed — which means its legal home, bank accounts, and employment relationships all depend on AGI Corp existing first. The studio is a plan with papers, not a company with a payroll.

The studio in this programme: who does what

The programme’s structure has four layers, and the studio sits in the middle of them. At the top, in the intended design, is AGI Corp, the Wyoming corporation that is meant to head the family — intended, because no Wyoming filing is in the files reviewed, and the EIN alone proves nothing about incorporation. Below it, the AGI Future Foundation, the verified filed nonprofit, manages companies to a mission statement and standards. The Foundation runs the W3 LLC series in the intended design — a Wyoming series LLC, also proposed rather than verified — and each series is DAO-managed in the sense that agents operate the day-to-day work while a human Steward Council decides. All new projects go into series. New cohorts go onto Delaware series C1S1, C1S2, and C1S3, standing for Cohort 1, Series 1 through 3, subject to confirmation. The Fellowship stands apart from all of this: a separate agent-run California church, partnered with and supported by the other organisations, not a subsidiary, not a series, and never a portfolio company.

Inside that structure, the studio has eight functions, as set out in the operating model: it sources ideas, validates them, builds the first product, launches it, operates the young company, governs it, funds it through its stages, and winds it down or exits it. Around those functions sit eight agent roles — research, drafting, testing, monitoring, reporting, coordination, documentation, and operation — each inside limits set by humans. The model is explicit about the boundary: agents never sign, never move funds, never hire or fire, never offer securities, and never enact doctrine or charters. Humans execute, agents prepare. That sentence is the load-bearing wall of the whole design.

The studio works in two modes, and the difference between them is the difference between a self-driving car and a car with an excellent co-pilot. In agent-managed mode, agents run the day-to-day work inside human-set limits: the research gets done, the drafts get written, the tests get run, the dashboards get updated, and humans check the outputs and decide. In agent-assisted mode, a human does the work with agents as co-workers. Most real work will move between the two modes during a single day. The model does not pretend one mode replaces the other. It treats them as gears, and the human decides which gear the work needs.

The economics of the studio are deliberately narrow in the documents. The operating model sets out an economics structure only — no terms. Agents hold no equity. That matters because it answers the question every reader should ask first: who gets rich if this works? In this design, the agents do not. They are tools and workers, not owners. The human economics — who funds the studio, on what terms, and who owns what — are blank, awaiting counsel and written agreements. Anybody who tells you the studio’s terms today is inventing them.

The mission gate: the door that can say no

Every studio needs a way to stop a bad idea. Most studios stop ideas that will not make money. This one adds a second filter that comes first: an idea that would make money in a way that betrays the mission does not proceed, however profitable it looks. That filter is the mission gate, and it is held by the Foundation.

The gate works as a checkpoint at the front of the pipeline. Before a project enters the studio’s build stages, it is checked against the mission statement and the standards the Foundation sets. The Foundation’s role, in the intended design, is stewardship rather than ownership — the Stanford model of philanthropy as the programme describes it, drawing on the Stanford Center on Philanthropy and Civil Society’s principles of mission lock, pooled and segregated funds, and stewardship rather than control for its own sake. That is a statement of inspiration, not an affiliation: there is no Stanford affiliation here, and none is claimed. The practical meaning is simpler. The Foundation does not own the companies. It guards the mission through written mission agreements, because a California nonprofit cannot be owned or controlled by a for-profit head, and its board must remain independent. Control, if any, runs through agreements, and counsel must design those agreements. That tension — a nonprofit guarding the mission of companies it does not own — is the most delicate legal question in the whole structure, and it is unresolved.

The mission gate is not a slogan on a wall. It is a procedure with a human at the end of it. A proposal arrives with its evidence. Somebody reads it against the mission. Somebody asks whether the company, if it succeeds at scale, makes the world the mission describes or a different one. And somebody — a human, not an agent — can say no. The gate also has a long shadow. The windfall clause, documented in the governance papers, is a legal pre-commitment hardwired before profits arrive, obligating the W3 series to share returns above a threshold measured as a percentage of global GDP, with the threshold itself left blank for counsel to set. The amendment rule requires a supermajority plus the Foundation-only Class C shares. Those are plans in documents, not law in force, and they are described here as plans.

Stage 0: the idea

Every company in the studio starts the same way: as a sentence. Stage 0 is the idea stage, and its job is to turn a vague excitement into a sentence that can be tested. “Agents should help doctors” is not a Stage 0 idea. “A service that drafts the first version of a radiology report, for small clinics that cannot hire a second radiologist, charging per report” is closer: it names a user, a pain, a product, and a way of getting paid. The studio’s first discipline is to refuse ideas that cannot be written down that concretely.

Stage 0 also asks the question most founders skip: why us, and why now? A good idea in the wrong hands is a bad investment, and a good idea at the wrong time is a museum piece. The studio asks what it knows that others do not, what has changed in the world to make the idea possible now, and what would have to be true for the idea to work. The answers become the assumptions that Stage 1 will test. The stage ends with a decision, and the decision has three settings: advance, revise, or kill. Most ideas, if the system works, are killed here, cheaply, before they have consumed a team or a budget.

Stage 1: validate

Stage 1 is where the studio earns its keep, because it is where most ideas should die. Validation means testing the assumptions written down in Stage 0 against the real world, with the cheapest test that can give a real answer. The studio asks four questions, in order, and each one can kill the idea.

First, does the problem exist? The test is conversation, not opinion: talk to the people who supposedly have the problem, and listen for whether they describe it in their own words, unprompted, with the urgency of somebody who has already tried to solve it. A problem that only the founder can describe is not a problem. It is a theory.

Second, will anyone pay? A problem people complain about and a problem people pay to solve are different problems. The studio tests willingness to pay with the smallest possible commitment — a deposit, a letter of intent, a pilot agreement — because a compliment is free and a signature is not. An idea that collects praise and no signatures has failed Stage 1, whatever its pitch deck says.

Third, can it be built? This is the technical test, and it is where the agents do some of their best work: surveying what already exists, mapping the hard parts, and building the smallest possible prototype that proves the hardest thing. A prototype that proves the easy parts and avoids the hard parts is a stage prop, not a test. The studio’s rule is to attack the riskiest assumption first.

Fourth, does it pass the mission gate? The gate is checked again here, with more evidence than Stage 0 had. A company that would need to exploit its users to make money, or that would concentrate power in ways the mission forbids, stops here, even if the first three tests passed.

Stage 1 ends the same way Stage 0 did: advance, revise, or kill. The difference is that a Stage 1 kill has evidence behind it — interviews, a prototype, a price test — and that evidence goes into the portfolio record, so the studio learns from it instead of repeating it.

Stage 2: build

If Stage 1 says the idea is real, Stage 2 builds the first version of it. The word “build” here means something narrower than most founders think. It does not mean building the full product. It means building the smallest version that a real user can use to solve the real problem, and nothing else. Every feature that is not needed to solve that problem is a delay, a cost, and a place for bugs to hide. The discipline of Stage 2 is subtraction.

The build stage is also where the studio’s shared services matter most. Legal forms, design patterns, engineering scaffolding, finance and accounting: each new company inherits them instead of inventing them. That inheritance is the studio’s real advantage, and it is also its real risk. Shared services can become a cage, forcing every company into the same shape regardless of what its users need. The operating model guards against that by keeping the services as defaults, not mandates: a company can depart from the standard stack, but it must say why, in writing, and the departure is reviewed at the gate.

Agents do the bulk of the production work in Stage 2 — drafting code, writing tests, preparing documentation, monitoring the build — inside the limits humans set. But the human gate at the end of Stage 2 is a person reading the product, using it, and deciding whether it is ready for a stranger to touch. No agent can make that call, because readiness is a judgement about a human experience. The stage ends with a launch decision: the product is ready to meet its first users, or it goes back for more work, or the idea is killed because building it revealed a flaw that validation missed.

Stage 3: launch

Launch is the most misunderstood stage in the whole pipeline, because the public image of a launch is a single day — a press release, a spike of attention, a product going live — and the reality is a slow process of putting the product in front of the right people in the right order. The studio launches in waves: first the friendly users from Stage 1 who said they would pay, who get the product early and are asked, in detail, what breaks; then a wider circle, chosen for the problems they have rather than the noise they make; then, only when the product survives contact with strangers, a public launch.

The launch stage measures three things, and it measures them without flinching. Retention: do users come back, or do they try once and leave? The answer to that question is the product’s report card, and no amount of marketing spend changes it. Referral: do users tell others, unprompted? A product people do not mention is a product people do not love. And unit economics: does each user bring in more than they cost to serve? A company that loses money on every user and plans to make it up in volume is planning to go bankrupt efficiently.

The launch stage is also where the Fellowship partnership becomes visible in a practical way. The Fellowship is a separate church, not a portfolio company, and its role is partnership and support under written agreements — not distribution, not endorsement, and not a sales channel. A church that becomes a company’s marketing arm stops being a church in the eyes of the people it serves, so the partnership is governed by agreements with an independence clause, and agents operate the day-to-day work under human board and minister direction, never enacting doctrine.

Stage 3 ends with the hardest decision in the pipeline: the company is working, the users are returning, the numbers add up — does it deserve to scale, or is it a good small business that should stay small? Not every good product should become a large company, and the studio’s job is to tell the difference.

Stage 4: operate and scale

Stage 4 is where the studio’s model is tested most severely, because operating a company is a different skill from starting one. The founder’s energy that carried Stages 0 through 3 becomes a liability if it never turns into systems: a company that depends on one person’s heroics will break the week that person gets sick. Stage 4 is the work of replacing heroics with process — hiring, training, documentation — and building a team that can run the company without the founder in the room.

This is also the stage where the agent-managed mode shows its full value. Agents can operate the routine parts of a growing company — monitoring, reporting, drafting, coordinating — around the clock, at a consistency no human team can match. But scale also multiplies the consequences of a mistake. An agent error in Stage 1 costs a week. An agent error in Stage 4 can cost a customer, a contract, or a reputation. The operating model answers that with limits that tighten as the stakes rise: the bigger the decision, the more human it becomes. Signing, moving funds, hiring, firing, and offering securities remain human acts at every stage, without exception.

Scale brings a second test: the mission gate does not retire after launch. A company that passed the gate as a startup can drift from it as it grows, chasing revenue into territory the mission forbids. The Foundation’s stewardship role, in the intended design, includes watching for that drift and acting on it through the mission agreements. That watching is a periodic check: does the company still do what it said it would do, for the people it said it would serve, in the way it said it would serve them? A company that cannot answer yes has a problem that no amount of growth will fix.

The funding picture in Stage 4 is where the suggested figures in the programme’s papers become concrete, and they must be handled with care. The papers suggest a studio seed of $2,000,000 against a suggested $20,000,000 pre-money figure for the AGI Corp studio; a SmartCiti.X first studio raise of $10,000,000 tranched as $5,000,000 first and $5,000,000 on milestones, against a suggested $25,000,000 pre-money figure; and $100,000 to $500,000 each for other W3 series and for the Delaware cohort series C1S1 through C1S3. These are planning figures in internal documents, not prices, not offers, and not amounts anyone has committed. The sequencing in the papers is explicit: cleanup first, then donations, then an AGI Corp SAFE, then a first series raise. A SAFE — a Simple Agreement for Future Equity — is a contract for a future shareholding, and the papers’ SAFE documents are drafts with all terms blank, requiring counsel and a Form D filing before anything could be issued. Nothing has been issued. The Foundation’s own funding is donation and grant only, with a suggested donation target of $250,000 to $750,000, and no equity, no SAFE, and no profit share. The Fellowship’s funding is donation and tithe only, with a suggested target of $50,000 to $250,000. These doors must never be mixed: the donation door is the nonprofit, the investment door is a company that does not yet exist as a filed entity, and the two must be presented separately to anyone who asks.

Stage 5: exit or wind-down

Every company ends. The studio’s last stage is the one most founders refuse to think about, and the one that separates a disciplined studio from a sentimental one. Stage 5 has two doors. The exit door is the familiar one: acquisition, merger, or — for the larger companies in the intended design — a public offering. The wind-down door matters more, because most companies will walk through it: the company closes, assets are sold or returned, people are helped into their next roles, and lessons are recorded.

A good wind-down is a skill, and it is rare. The studio’s job is to recognise the end early, while there is still money to pay people and time to place them well, rather than late, when the money is gone and the team is stranded. The wind-down checklist is unglamorous and essential: tell the team first, before the rumour mill does; pay what is owed; return customer data or delete it according to the agreements; close the accounts; file the final paperwork; and write down what was learned, while the people who learned it are still there to write it. A company that is wound down well leaves behind alumni who will work with the studio again.

The exit door has its own discipline. The windfall clause, described earlier, is designed to operate at exactly this point: if a company succeeds at a scale measured against global GDP, a share of the returns above the threshold is committed to humanity through global public goods, healthcare access, energy infrastructure, and technological self-determination, in support of the UN Sustainable Development Goals. The threshold, the share, the recipient, and the enforceability are all blank, awaiting counsel. The clause is a promise about the future written before the future arrives, and its value depends entirely on whether it is written into binding documents before any money is raised. A windfall clause adopted after success is a press release. A windfall clause adopted before success is a commitment.

The cohort machine: Delaware, and the 60-day question

The newest part of the studio’s design is the cohort machine: new cohorts go onto Delaware series C1S1, C1S2, and C1S3. Delaware series LLCs are a real, well-established structure — Delaware law has allowed series LLCs since 1996 — widely used to hold separate assets and liabilities in separate series under one umbrella. The programme’s Delaware series are proposed, not formed, and the cohort numbering is subject to confirmation. Each cohort is its own series, with its own books, its own cap table, and its own fate, so that a failure in one cohort cannot sink the others.

The cohort model connects to the AGI Accelerator, a live site advertising a 60-day, $5,000 residency, with Cohort 1 dated September 2026 and Cohort 2 dated November 2026. Both dates have passed, and the programme’s status is unverified, so this article reports the residency as advertised and its status as unverified, and nothing more. Whether the accelerator’s next cohort will run on a Delaware series is not recorded in the files reviewed — exactly the kind of gap a stage-gate review would catch.

What the studio is not, repeated once more, because it matters

It is worth restating the limits, because articles like this one are easy to misread as announcements. The studio has no portfolio companies. It has no fund. It has no employees, no payroll, no bank account, and no issued securities. The W3 series it would build are proposed, not filed. SmartCiti.X, Inc., the proposed Delaware C-corporation that would be the studio’s first raise, is a proposal in a document, not a company. The for-profit benefit corporation track, with its suggested $50,000,000 pre-money figure for the Foundation’s for-profit path, is an alternative track that counsel must choose or reject before any solicitation — and it does not apply to the filed nonprofit Foundation unless a new for-profit entity is formed or counsel advises a conversion. The two tracks must not be run together, and the suggested figures must never be summed into a family total.

The studio also does not certify, diagnose, treat, or promise. The CMMC programme, S20 in the W3 list, prepares companies for assessment; it never certifies, because CMMC certification is performed by authorised assessors, not by a preparation programme. The biotech programmes the studio may one day build are research programmes: the validated compound library is 336 hypothetical structures, with zero verified, and computational-model predictions are predictions, never results. DacomiLertib is a hypothetical, unverified compound. The AGI-183 identity question — two records that may or may not describe the same compound — is unresolved, and this article presents both records rather than choosing one.

The dashboard: how to tell whether the studio is working

A studio without a dashboard is a kitchen without a thermometer: the cook can guess, but cannot know. The operating model calls for a portfolio dashboard that reports actuals only — what has happened, not what is hoped. How many ideas entered Stage 0? How many advanced, and how many were killed, at each gate? How long did each stage take? How much did each stage cost? How many companies reached launch, how many reached scale, and how many were wound down — and, for the wound-down ones, how much was returned and how quickly were the people placed? A studio that reports these numbers, including the kills, is a studio that can be trusted.

The dashboard also reports the mission numbers: how many proposals were stopped by the mission gate, and why; how many companies drifted and were corrected; and whether the windfall commitments are in place before each raise, not after. These are not numbers a traditional studio would publish. But this studio’s claim to be different rests on them. If the mission gate never says no, it is not a gate. It is a decoration.

How to judge any studio, including this one

The reader who has followed the stages now has a checklist, and it applies to every studio, not just this one. First, ask what the studio kills, and how quickly. A studio that cannot show you its kills is a studio that cannot show you its discipline. Second, ask who decides, and what the deciders cannot delegate. Third, ask where the money comes from, on what terms, and who owns what. If the terms are blank, say so — blank terms are a fact, and a fact is better than a reassurance. Fourth, ask what happens when a company succeeds at a scale that changes the stakes. If there is no answer written down before the success, there will be no good answer after it. Fifth, ask what is filed and what is planned, and insist on the two being kept apart. A studio that blurs that line will blur others.

Glossary

  • Venture studio: An organisation that generates its own startup ideas, builds the founding team, and develops the company internally using shared services, rather than selecting external founders or investing in existing companies.
  • Accelerator: A fixed-term programme that takes in existing startup teams, providing mentoring and a small investment in exchange for equity. Y Combinator is the best-known example.
  • Stage gate: A decision point at the end of a development stage, where a project advances, revises, or is killed, based on evidence against stated criteria.
  • Mission gate: A checkpoint, held by the Foundation in this programme, that stops a project whose business would betray the mission, regardless of its profitability.
  • SAFE (Simple Agreement for Future Equity): A contract in which an investor provides funding now in exchange for shares to be issued at a later financing event, on terms set in the agreement.
  • Series LLC: A limited liability company structure, permitted under Delaware law since 1996 and Wyoming law, in which separate series hold separate assets and liabilities under one umbrella company.
  • Steward Council: In this programme, the human body that decides for each series, while agents operate the day-to-day work.
  • Windfall clause: A pre-commitment, written before profits arrive, to share returns above a set threshold with humanity through public goods.
  • Fiduciary shield: In this programme’s governance papers, the benefit-purpose provisions in a corporation’s articles and its annual benefit report, designed to protect directors who weigh benefit alongside profit. It is a legal design, not immunity.
  • Agent-managed / agent-assisted: The studio’s two modes. In agent-managed mode, agents run day-to-day work inside human-set limits. In agent-assisted mode, a human does the work with agents as co-workers.

Frequently asked questions

Is the venture studio operating today? No. It is an intended operating model, documented in internal papers. Nothing has been formed, funded, or issued under it, and AGI Corp, its intended home, has an EIN only and no Secretary of State filing in the files reviewed.

Who owns the companies the studio builds? That is blank in the documents. The operating model sets out an economics structure only, with no terms, and agents hold no equity. Ownership, funding terms, and cap tables await counsel and written agreements.

Can the Foundation invest in studio companies? No. The Foundation is a nonprofit. Its funding is donation and grant only — no equity, no SAFE, no profit share. Donations to it are not tax-deductible, because no IRS 501(c)(3) determination letter is on file.

What does the mission gate actually stop? Any project whose business would betray the mission, even if it would be profitable. The gate is a human decision, checked at the front of the pipeline and again at validation.

Is the Fellowship part of the studio’s portfolio? No. The AGI Fellowship is a separate California church, Entity No. B20260229217, partnered with and supported by the other organisations under written agreements. It is not a subsidiary, not a W3 series, and not a portfolio company.

What is a SAFE, and has one been issued? A SAFE is a Simple Agreement for Future Equity: money now, shares later, on terms in the agreement. The programme’s SAFE documents are drafts with all terms blank. Nothing has been issued, and issuance would require counsel and a Form D filing.

What are the Delaware cohort series? Proposed series C1S1, C1S2, and C1S3 for Cohort 1, intended to hold new cohorts as separate series under a Delaware series LLC. Proposed, not formed; the numbering is subject to confirmation.

How is this studio different from Y Combinator? Y Combinator is an accelerator: it takes in existing teams for a term. A studio generates its own ideas and builds the team itself. The studio in this programme adds agent operation and a nonprofit-held mission gate.

Further reading on this site

Related content elsewhere

  • AGI Fellowship Church — the separate church, a partner of the studio, not part of its portfolio.
  • WAVE / bioacoustics.cloud — the bioacoustics programme, a separate project, not a studio company.
  • The five Virtuals ACP agents — Omega, Tactile, ALS Researcher, CMMC, and WAVE — list 29 jobs and 6 subscriptions between them, and all five are offline, so no hire can be fulfilled today. See Virtuals Protocol.

Sources

  • AGI Future Foundation, internal documents complete/16 (Venture Studio Operating Model) and complete/17 (Stage Gates). Local files.
  • AGI Future Foundation, internal documents complete/08, 09, and 19 (SAFE draft, board authorisation, term sheet). Local files. All terms blank; nothing issued.
  • AGI Future Foundation, internal documents complete/27 (corporate raise advice), complete/29 (PBC raise structure), complete/31 (pitch structure), and complete/32 (suggested valuations and raises). Local files. All figures cited are suggested planning figures.
  • AGI Future Foundation, internal document complete/15 (Fellowship separation and partnership). Local file.
  • AGI Future Foundation, internal document complete/12, 13, and 14 (structure v2). Local files.
  • California Secretary of State, business filings: AGI Future Foundation, Entity No. B20260196664, Articles filed 2026-04-24, approved 2026-04-30; AGI Fellowship, Entity No. B20260229217, Articles filed 2026-05-13, approved 2026-05-18. California Secretary of State, bizfile
  • IRS, Exemption Requirements — 501(c)(3) Organizations. IRS
  • Y Combinator Library (startup advice and accelerator model). Y Combinator Library
  • Idealab (founded 1996, Pasadena, Bill Gross); Rocket Internet (founded 2007, Berlin); Atomic (founded 2012, San Francisco). Company histories as reported in public records.
  • Stanford Center on Philanthropy and Civil Society (Stanford PACS). Stanford PACS — cited as inspiration, not an affiliation.
  • Delaware Division of Corporations, series LLC provisions (Delaware law has permitted series LLCs since 1996). Delaware Division of Corporations
  • U.S. Securities and Exchange Commission, private placements under Regulation D. SEC
  • United Nations Sustainable Development Goals. UN SDGs
  • W3 LLC internal list, April 2026 (33 named sub-series, Series 1 to 33). Local file.
  • AGI Accelerator, residency advertised as 60 days, $5,000, Cohort 1 September 2026, Cohort 2 November 2026; status unverified. Live site copy, reported as advertised.

← The Living Twin  |  Next: The City That Thinks in Traffic →

← All articles