A system that validates itself
I recently came across a concept from philosophy that keeps showing up in different parts of my life: self-referentiality. A system is self-referential when it produces its own criteria for validation. Instead of being measured against something outside itself, it measures itself against itself.
This is not a flaw by default. Language works this way (every word in a dictionary is defined by other words), and so does money, which is valuable because it is accepted and accepted because it is valuable. But some systems turn self-reference into a business model. Two worlds I have spent more time in lately share this trait: art and venture capital.
One of the questions I’ve been exploring since is: if a system validates itself, what happens to the people who build outside of it?
Art: the circuit that makes you big
When an artist is chosen for a show at MoMA, they have joined a circuit in which the mere fact of belonging will make them big. Being at the exhibition does not only recognize value; it produces it.
The circuit runs on people with credentials: curators, critics, gallerists, and collectors authorized to say what deserves a museum wall. Philosopher Arthur Danto called this the artworld, the atmosphere of institutions and theory that turns an object into art. Bourdieu went further: the real question is not who creates the work, but who creates the creator.
The clearest expression of this is provenance. The value of an artwork includes where it has been. A painting that once belonged to a great collector is worth more than an identical one that did not, because its history of ownership has become part of the work.
Venture: kingmaking and the cap table as provenance
Venture works in a similar way. The credentialed players are not neutral observers. They issue prophecies (“this is the next big category”) and they have financial incentives for those prophecies to come true. Their credential is precisely what raises the odds that they will.
To be fair, top funds also see the best deal flow and pick well; much of their companies’ outperformance would have happened anyway. Still, part of it is signaling. When a top-tier fund leads a round, investors, candidates and customers infer quality from the brand, so the next round fills faster and prices higher. The company does become more valuable, partly because everyone believes it will. The industry calls this kingmaking.
The cap table is venture’s provenance. When we describe a deal, the names of the investors who came before are not decoration; they are information about the odds that the prophecy will be fulfilled. A company is worth what it is worth because enough people believe it is worth that, act accordingly, and by acting make it so.
The circle closes
There is nothing wrong with a system working like this. This is not a hidden dynamic. YC publishes Requests for Startups, top funds share their big ideas for the year, and many founders build toward them. I believe those lists are genuinely useful: they help talent and capital find each other. But the clearer the signal, the more people build toward it. If artists follow what curators announce and founders follow what top funds publish, the circle can start to close, and the system begins to confirm its own predictions.
What disappears is the classic leap into the unknown: the founder or artist who jumps without a guarantee, against the consensus, with no one to catch them if they fall. In a perfectly self-referential system there is no outside left to jump into, only bets with a safety net.
One of venture’s oldest rules of thumb captures the problem: to achieve superior results, you have to be non-consensus and right. But if consensus backed by enough capital makes you right, being non-consensus stops being an epistemic virtue and becomes a timing error. The system punishes the bold bet in the short run, and only pays it once the circle catches up.
Elaboration and consecration
And yet the circle never fully closes. A Defiance Capital study of 845 US and UK unicorns found that only 28% had a top seed fund in their seed round, and even Sequoia got into fewer than 3% of them. At the start, the winners are scattered and most of the system misses them. Yet a separate analysis by Affinity found that about 20 firms appear in nearly half of all unicorns, which suggests the big names often arrive once the company is already visible. The circle does not so much create its winners as crown them.
This has changed my mind about where founders need to be. As someone from Latin America, with Korean roots, now based in New York, I have always lived between centers and peripheries. For a long time I believed that a founder outside the US had to move to San Francisco as soon as possible to play in the big leagues. Then, with AI, I thought it was not necessary, at least for some industries. Now I think it depends on two different moments in the creation of value.
The first is elaboration: building the product, doing the science, writing the code. The second is consecration: the moment the work is lifted up and recognized by those with the “authority” to do so.
If you want to sculpt in marble, you go to Italy, because the stone and the workshops are there. If you write poetry, you can do it anywhere. Elaboration, for most founders today, is poetry: it can happen anywhere, and often happens better far from the noise. Hard tech is closer to marble: the material decides where the work happens.
Consecration has centers, the way Rome once did for sculptors and Paris did for painters. For AI and software, the center is still mostly San Francisco. But consecration is starting to fragment by vertical. El Segundo has turned the SpaceX diaspora into a cluster of defense, space and energy hardware startups. Munich has become Europe’s defense tech hub around Helsing and Quantum Systems. Boston has long played this role for biotech, Tel Aviv for cybersecurity and São Paulo for Latin American fintech.
Notice where these centers form: near the customer and near the matter (a military command, an aerospace legacy). In deep tech, part of the consecration comes from those who buy and test, not only from those who invest. In these sectors, consecration gets pulled toward the places where elaboration happens. So the question is not whether founders need to be in San Francisco, but which center matters for what they build, and when. Build where you have an edge; show up at the right center when it is time to be lifted.
Consecration before elaboration
In September, Andreessen Horowitz announced the Horowitz Andreessen Academy: a tuition-free, in-person year in San Francisco for about 50 recent high school graduates, starting in September 2027, backed by $42 million led by a16z. Admission is based on proof of work (what applicants have built or shipped) rather than grades or essays. Founding partners include Anthropic, OpenAI, Nvidia and Stripe, and the firm has said it wants both stakes in the startups that come out of it and a pipeline of graduates into its portfolio companies.
In a way, it is a deeper version of the annual reports in which top funds announce next year’s hot sectors: instead of predicting the trend, it trains the people who will fulfill it. The generous reading is that this opens the circuit. Proof of work instead of credentials is exactly how you would design a door for the best founders, including the ones the circuit would otherwise miss.
The less generous reading is that it closes the circle even further, by moving consecration earlier than elaboration. The art world already tried this. In 19th century France, the Académie trained the painters and the Salon consecrated them. In 1863 the Salon rejected so many works that the emperor ordered a Salon des Refusés to show them. A decade later, the outsiders stopped waiting for permission, organized their own exhibition, and became the Impressionists.
An academy owned by the kingmaker does not wait for the gem to appear; it consecrates at 18 years old, in San Francisco, before anyone else has had the chance to look. It also bets against my own change of mind: it assumes that even elaboration has to happen at the center.
Where matter pushes back
Consecrating before anything is built works best in software, where a narrative can sustain a valuation for years. It works worst where matter pushes back. In deep tech and critical infrastructure, the reactor works or it doesn’t, the material holds or it doesn’t, the grid delivers the megawatts or it doesn’t. Capital can inflate a valuation, but it cannot inflate a substation.
A mining company does not care who led the seed round. It cares whether the process works. In these sectors the bold bet still has a chance, because the system cannot decide on its own whether it was right.
What this means for allocators
Every system is self-referential to some degree, so the useful question is not whether venture validates itself. The question is how much it can still be disturbed by things outside its own code: a customer who pays, a law of physics, a founder without a pedigree.
That gives early-stage funds, especially those far from the center, a clearer role. How do you find the winners most of the system misses at seed? Look for founders doing serious elaboration in Buenos Aires, Barcelona or a lab nobody in San Francisco is watching, often in sectors where matter keeps everyone honest.
If the center is now recruiting right after high school, the local fund’s edge is proximity to elaboration, and to the customers who test it. But proximity alone is not enough. The circle runs on signals: who invested, which thesis it fits, what the last round priced. The outsider can only be judged on substance, and that takes human rigor: understanding the niche, the market’s context, the technical risk, and the founder up close. That work does not scale, which is exactly why the circle skips it. The fund’s job is to find the gem before it is extracted, and then build the bridge that brings it to the center at the right moment, on better terms.
At 1200vc, we see this from both sides. We partner with early-stage funds and also invest directly, and that is exactly the challenge when we look at opportunities: telling real substance from a strong signal. We believe our fund managers are the kind of players this essay describes, investors who understand the nuances of their market, their sector and their founders up close. That is what lets them find the companies the circle has not seen yet, and what lets us back them before it does.
I am not outside this circle either. I work in venture, and an essay like this one is a small bid for consecration of its own. That is exactly why the question matters to me.
Here is what I keep coming back to:
In venture, returns do not come from being right. They come from being right before the circle agrees. The gap between elaboration and consecration is where the money is made.
Every move that pulls consecration earlier (a signal round, a published thesis, an academy that crowns at 18 years old) narrows that gap for everyone inside the circle. So the more self-referential venture becomes, the more its returns depend on what it cannot yet see. The circle needs its outsiders more than they need it: a system that only validates itself eventually runs out of anything new to crown.
The edge is not in breaking the circle, but in standing where it has not arrived yet.
I’m still thinking this through. What happens to the people who build outside the circle? I’d love to hear how you see it.
Micaela Kim
October 2026
