AGI is the energy.
Human collectives are the liquidity.
Persist is the fulcrum.

Persist is a talent activation and placement machine — a venture fellowship that finds untapped builders, pairs them with credentialed co-founders, and points them at the frontiers that matter. This is the thesis, the model, and how we’re evolving.

2.0  Capital efficiency & credibility Next  Mycelial capital allocation Clock  Weakly general AI, Nov 2027
$2M → $115M
Raised, converted to
projected NAV since inception
345K+
Engineers funnelled into
the ecosystem, internationally
60B+
Views generated for clients
and internal projects
40
Portfolio companies
from a single raise

01  /  Origin

A rejection letter that turned out to be a business model.

The Thiel Fellowship gives $100,000 to a handful of young people who leave college to build. It works — spectacularly. It is also a charity, which means it captures none of the value it creates, and turns away tens of thousands of applicants who were hungry to prove they belonged.

One of them was our founder. That denial produced the calculation the entire company is built on.

$22M
Total grants deployed
220 fellows × $100K
$220B
Collective portfolio value
generated by those fellows
$166.7M
Average value created
per fellow-year of labour
$80,128
Value created per fellow
per working hour
20
Selected per year
out of tens of thousands

Grant total is 220 fellows × $100K. The $220B is the collectively reported net worth of Thiel Fellowship alumni (Business Insider, 2022: “power players worth more than $220 billion collectively”) — a figure for the people, not an audited attribution of value to the programme. The per-hour number divides that total across the cumulative fellow-years the cohorts have been running, then by a 2,080-hour year. It is a back-of-envelope order-of-magnitude argument and is used here only as one: the conclusion survives being wrong by a factor of ten.

The first-principles version

If there were economic value in paying someone minimum wage to work on their own big vision — even if they only owned 1% of it — that would beat any other job available to them:

  1. You are your own boss, working on something that matters.
  2. You take the same or better direct salary than the alternative.
  3. You hold ownership in the thing you are building.
  4. You work at the frontier of your own potential for impact.

Run the Thiel numbers and the conclusion is unavoidable: if even 10% of the value created accrued to whoever funded that life path, it would be extraordinarily worth doing. No such option existed. That gap is the company.

Access to capital for the career path of innovation is not de-risked at the earliest stage — the one stage where it should look like a job market.

The founding insight
The proof case

Denied, then validated anyway

The year of the rejection, our founder built the first PayPal–to–Ethereum exchange, because none existed. Within four months of launch it had processed over $500,000 at a 20% margin — the only provider in the 2017 market, organic, with zero marketing spend.

Then a single large fraudster wiped out the entire profit, and there was no capital left to continue.

The lesson

The product wasn’t missing. The organisation was.

An ecosystem would have recognised early product-market fit, raised millions against it, and lifted the buy limits. Instead: no connections, no oversight, no follow-on — and a working business died of a solvable problem.

Automation, mentors, recruiting, legal, distribution, capital. The value of a head organisation sitting above the venture is immeasurable, and it was the one thing not available at any price.

02  /  Thesis

The Directional Economy.

Intelligence is becoming abundant. Direction is not. Persist’s bet is that the scarce asset in an AGI economy is not cognition — it’s the ability to aim cognition at something that matters, and to keep humans inside the loop that decides what does.

I

The orchards of the future

A new class of orchards is emerging — fields of opportunity abundant with low-hanging fruit, appearing wherever intelligence, computation and coordination create asymmetric returns. The pickers are not farmhands. They are technically literate people distributed across the world, newly equipped with tools that multiply what they can produce.

Persist’s job is to direct that potential at the most fertile terrain faster than anyone else — not through control, but through directionality: detecting anomalies in human capability and aligning them toward frontier-scale objectives.

II

The honest counterargument

If intelligence becomes abundant, the bottleneck moves to application — and the largest threat to this entire thesis is that AGI captures the majority of the potential it creates, becoming a fully autonomous venture studio with little human element required. An AGI that internalises the venture loop leaves no inefficiency for humans to exploit.

We take that seriously rather than waving it away. It only happens if AGI can reach the full informational substrate of reality — including the cultural, emotional and narrative dimensions that actually govern behaviour.

III

Where a static model fails: liquid markets beat past experts

As Polymarket’s founder frames it, a neural network is a past expert trained on the market of ideas. It aggregates a training corpus — prior ideas, weighted by co-occurrence and gradients — but it is a static snapshot. No real-time bid–ask. No incentive-driven updating. No adversarial participation. A true liquid market requires ongoing human stakes: reputation, capital, emotion.

Humanity is that live market. Which is why humans plus AI beats AI alone, and why the defensible position is not being smarter than AGI — it’s building the liquid markets for intelligence that AGI has to learn from.

IV

Resonance data is the next scarce substrate

Views and likes are binary. They record that something worked and lose why — which makes success very hard to reproduce. The next layer of feedback comes from richer, participatory signal: 0–100 ratings, facial-expression response, heart-rate change, tone. Call it resonance data.

That data cannot be scraped off existing platforms. It requires new platforms built to collect it. Whoever builds them controls the most valuable training substrate of the century — and those platforms are collectively operated cultural engines, which is exactly the kind of thing a fellowship of builders is positioned to launch in parallel.

V

Super-persuasion > super-intelligence

A super-intelligence can model everything that has been. Only a super-persuasion system can steer what becomes. What is the value of intelligence, other than the ability to persuade reality into the shape you want?

In the long run the most valuable companies will not be the ones that possess intelligence. They will be the ones that direct it — that convert cognition into coordination.

VI

The fulcrum

Archimedes: give me a lever long enough and a place to stand, and I will move the world. AGI is the lever — the raw force. The place to stand is direction.

Persist becomes a collective intelligence refinery: feeding frontier systems human novelty, while extracting meta-insight from how those humans actually succeed. Over time, the onboarding layer of the AGI economy — teaching machines what humans value, and teaching humans how to wield machines.

It doesn’t compete with AGI. It completes it.

The fulcrum era
Read the plain-language version

1. The world just changed

The internet used to be about sharing information. Now it’s about sharing intelligence. AI can code, design, write, even start companies. The amount of smartness in the world just exploded. But intelligence without direction is a rocket with no target — powerful, and burning fuel going nowhere.

2. What Persist actually does

Imagine billions of new orchards of opportunity suddenly appear. The problem isn’t the fruit — it’s that most people don’t know which direction to walk. Persist is the map: it shows skilled people where to go, which fruit to pick, and how to team up to reach the next orchard faster.

3. The big threat

Some worry AI will soon start its own companies and leave humans behind. There’s one thing it still can’t do well: understand culture. It can predict what worked before. It can’t fully understand why people care — what makes something funny, moving, or meaningful.

4. Why humans still matter

AI is a super-expert trained on everything that already happened. Humans are a live market — always changing, reacting in real time. The best results come from both: humans bring emotion, intuition and curiosity; AI brings speed, memory and focus.

5. The secret ingredient is direction

The future won’t be won by whoever has the smartest AI. It’ll be won by whoever can aim it. Persist finds the people with the spark and gives them the tools, partners and feedback loops to turn ideas into world-changing projects.

6. In short

AGI is the energy. Humans are the heart. Persist is the lever that moves them both in the same direction.

03  /  Model

Venture fellowship, not venture studio.

The rename is not cosmetic. Venture studios carry a reputation problem in the investment world: one team building several projects at once, which usually means a single unfocused founder who hasn’t found conviction in anything. The studios that actually work run like YC — a machine that takes capital in, returns more value out, and reinvests in its own efficacy.

What we run is closer to a Thiel Fellowship with a cap table. Hence: venture fellowship.

Thiel FellowshipTraditional pre-seedAccelerator (YC-type)Persist
Who it targetsDreamers and visionaries, pre-alignmentFounders already inside the networkTeams that have already formedDark, untapped talent from the job market — pre-founder
Capital per venture$100K grantA priced round, typically seven figuresStandard cheque + programmeRoughly 10× less per startup, spent as needed
Ownership takenNone — it’s a charityA minority stake for the fundA small fixed percentageRoughly 10× more equity per venture
Founder incentiveGrant, then you’re on your ownEquity-ledEquity-ledSalary-led out of the gate, 3–5× raise post-funding, equity taught as the real prize
Selection basis~20 chosen per yearWarm intro and pedigreeBatch interviewOpen competitive funnels at scale, judged on shipped work
Value captureZero by designFund carryProgramme equityCompany value, not just fee or SaaS value
Failure of the modelTurns away the restBets big, waits yearsDepends on inbound qualityCap tables that look unusual to follow-on investors

The Persist column is drawn from our own numbers. The other three columns are qualitative characterisations of how those models are generally understood to work, offered for orientation — not measured comparisons.

Invest pre-founder

The outsized return comes from entering earlier than the category allows. We don’t find founders — we find capability, then help the person step into the founder archetype: craft the story, build the supporting materials, and put an ecosystem behind them from day one.

A positive side-effect of high ownership: a project whose founder isn’t working out can be re-led without losing the investment in the product. And it leaves room to bring experienced executives in later on genuinely motivating equity.

The arbitrage stack

10×
Less capital
deployed per startup
10×
More equity retained
per venture backed
6×
Larger STEM talent pool
in the markets we hunt
15×
More labour value
per dollar of salary
30×
Less competing
VC capital in that market

Comparisons are against the standard US accelerator baseline. Each multiplier compounds the others: cheaper shots, more ownership per shot, more shots available, less competition for the talent taking them.

The correction

We over-indexed on technical skill. We under-indexed on credentials.

Venture funding is won on social capital more than on team merit. Harvard, MIT and Stanford are the powerhouses of that capital: their students raise $3–4M seed rounds on average, and roughly 25% of all seed-stage dollars flow to those schools — on the order of $480K of seed capital per student, per graduating class.

Meanwhile we have supplied several venture-backed startups with engineers from our pipeline who outperform their $300K-a-year Bay Area counterparts. Both facts are true. Only one of them raises money.

The fix

Pair the credential with the cracked operator.

The archetype in full form is two people: one focused on network, narrative and evangelism; one who ships relentlessly on the insight. The 1-in-10,000 developer — full-stack, real design taste, fast learner, faster shipper — is already funded and co-founding by graduation if they’re at Stanford. Elsewhere, they lie dormant.

Early evidence  ·  50+ calls with Stanford students, first placements made, three portfolio companies matched with credentialed co-founders carrying real celebrity and operator networks.

The structural trade-off, stated plainly

What the model does to a cap table

Persist replaces the standard 50/50 CEO–CTO split with a holding entity: the technical founder holds a meaningful minority of it, Persist holds the majority, and the entity occupies the co-founder seat. Combined with a salary-first incentive, that leaves the operating founder in the 10–20% range at the point where an early-stage investor first looks at the table.

Why that is a real objection

Follow-on investors who care about founder ownership are not used to seeing it that low that early. It is the single most legitimate criticism of the model, and pretending otherwise would be a mistake.

The three things that resolve it

A partner fund or side-car vehicle that normalises the structure; one large success case that makes the structure irrelevant; and a willingness to move toward lighter ownership where it unblocks a raise. Even at materially lower ownership, the capital efficiency of the model still holds — the arbitrage is in the entry point, not the percentage.

The test that actually matters

A studio only earns the right to this structure by proving it deploys received capital better than the standard selection basis does. That is the bar we hold ourselves to, and the reason the first phase was spent proving technical delivery before anything else.

04  /  Machine

Funnels plus automations, converted into outcomes.

Persist 2.0 is, operationally, one sentence: become more capital-efficient at turning funnels and automations into outcomes. Every funnel below lowers burn per projected outcome, widens the pipeline of connectivity to experts, and anchors onboarding to a culture of high expectation.

Funnel 01

Startupathon

Open competitive challenges against real listed projects. Builders self-select in, ship against a brief, and the work is the application. The message is deliberately filtering: don’t join unless you’re ready for a growth mindset.

Captures volume at the top — and the people who don’t win are routed onward rather than lost.

Funnel 02

Co-Foundathon

Matching, not recruiting. The mechanism that pairs a credentialed co-founder with a cracked technical lead around a specific thesis — the funnel that directly addresses the credential gap.

Captures the missing half of every founding team.

Funnel 03

Accelerator

Build with us, keep 93%. Seven percent equity, no cash: credits, courses, materials, resources, automations, and a kit. Pure upside for both sides, and the cheapest possible way to widen the top of the funnel.

Captures founders who don’t need money — only leverage.

Funnel 04

Request for Startups

An RFP platform pointed at the other side of the market. Someone with foresight, domain expertise or market insight posts the idea, the requirements, the funding and the equity they’ll part with — and waits for the world to build it.

Captures ideas from people with no time, and turns the world into bounty hunters.

Funnel 05

Createathon & Jobathon

Two narrower nets. Createathon hunts creators who have already proven distribution instinct: go semi-viral once and you are very likely hired, twice and it is a certainty. Jobathon keeps every recruited cohort able to compete for the next role, with completed challenges as public social proof.

Captures distribution talent and keeps the army warm.

Funnel 06

Residency & campus programmes

Three live experiments aimed squarely at credentialed co-founder placement, run directly into elite campuses rather than waiting for inbound.

Captures the social capital that the venture market actually prices.

You don’t die by failing. You die by going too slow.

The operating standard
Layer 02

The distribution stack

Most companies raise money to buy attention. It is a considerably better angle to raise attention in order to buy money. Distribution is the life force that can make nearly anything come alive — so it gets treated as an in-house discipline with named owners, not an afterthought once the product ships.

GIF SEO — best in the world Micro-influencer strategy Content posting operations Mass cold email LLM SEO Google SEO ASO

Because we hold equity rather than sell software, each of these tools compounds into company value, not SaaS value — which is why we can afford to be the best at them.

Layer 03

Onboarding as identity transformation

The single most underrated lever is set and setting. Receiving the letter should feel like walking through a portal — an offer that reads as a difficult quest with a tour of duty attached, not an employment contract. Only sign it if you’re ready to be pushed.

Then the status loop: public daily updates to a real network rather than a private tracker, kit that people are proud to show, and an acceptance moment worth sharing. Once a founder is updating their own network and not just us, a different level of care switches on.

Governing belief  ·  Mental ownership is infinite. Make sure the founder owns their company in their head at 10,000%.

The failure modes we are actively fixing

Great products, no growth budget

We consistently assembled strong technical founding teams and reached good initial product. Then the absence of marketing capital at exactly that moment capped every user-growth attempt. Fix: an in-house growth consultancy that understands the full stack and tests for a profitable funnel alongside each company, plus one senior storyteller verticalising distribution across the whole ecosystem.

Founder knowledge is the bottleneck

From inside a project, progress feels fine. From outside, it is often simply too slow — and slow is fatal, because business is a race to market and the frontier keeps moving. Fix: real oversight on product and UX, structured curriculum that teaches thinking rather than tasks, and mandatory use of co-founder and mentor search rather than waiting to be introduced.

The ecosystem was siloed

An ecosystem this large should be hyper-connected to the outside world and is not yet. Fix: advisors and mentors holding real equity in sub-companies, demo days at 1, 3 and 6 months for subscribing firms, and the portfolio page repositioned as an advisor opportunity surface.

Daily updates went corporate

Salaries attract people who value doing the work and collecting the cheque — which is the opposite of the hunger the model needs. If someone cares more about their job than their company, we have an employee, not a founder. Fix: updates go public, status aligns with company success, and support scales with demonstrated obsession rather than being distributed evenly.

Too much engineering, not enough marketing

The correction is not less engineering — it’s engineering pointed at distribution: outreach tooling, application automation, conversion testing, talent-spotting systems that surface the young and hungry before anyone else notices them.

Leadership concentration

Strategy has been too dependent on one mind and too thinly propagated. Fix: senior leadership that owns domains jointly and holds each other accountable, and a dedicated function on collective intelligence — turning daily updates and activity data into an actual read on who is moving and who is performing motion.

05  /  Evolution

1.0 proved delivery. 2.0 buys credibility. 3.0 changes how capital moves.

Each phase corrects a specific, identified failure of the one before it. Nothing here is a pivot — it is the same machine, with the constraint moved.

Phase 1.0  ·  Complete

Prove technical prowess

Hunt for wicked builders at scale, and find out how many startups you can nurture by not investing everything up front. We know within three months — sometimes one — whether a project is worth continuing, which makes burning as needed hyper-efficient.

The result: 40 portfolio companies and an ecosystem of over 200 people, with products brought to market on a single raise — drawing on a worldwide network of more than a thousand contributors.

What it taught usWe over-indexed on technical skill as the thing that raises venture capital, and under-indexed on credentials. Both halves are needed. Only one was being hunted.
Phase 2.0  ·  Now

Capital efficiency & the head of the snake

The core insight of this phase is uncomfortable: the overwhelming majority of value comes from the head — from selling what the rest of the entity is getting done — and the head has not had enough attention.

So: an overarching senior team that owns domains jointly; a materially higher bar on devotion and attention to detail; far heavier reliance on funnel data and statistics; and new funnels that cut burn per projected outcome while widening access to experts.

The unlockRising credibility compounds into everything else — the ability to tell high-calibre people what to focus on with their life, and to strike partnerships that were previously out of reach.
Phase 3.0  ·  Next

Mycelial capital allocation

Traditional venture bets big on few seeds and waits years. A fungal network does the opposite: it sends filaments out in every direction, reinforces the strands that find food, prunes the rest, and redistributes nutrients across the whole organism.

Translated into capital: scatter small exploratory cheques like spores, sense where cultural traction is actually metabolising, channel resources hard into the shoots with exponential uptake, and cross-link the portfolio so one collapse doesn’t poison the network.

Why it fits usWe already run wide exploration. This phase makes the reinforcement step deliberate, measured and fast — a living adaptive allocator rather than a rigid fund structure.
1 2 3 4 5 Spore release Hyphal probing Gradient sensing Reinforcement Redistribution RETURNS FUND THE NETWORK, NOT JUST THE FUND
01Spore releaseDeploy small, cheap, exploratory cheques widely — spores landing everywhere rather than one concentrated bet.
02Hyphal probingTeams use those cheques to run fast experiments directly in the cultural substrate: campaigns, proto-apps, launches.
03Gradient sensingRead where traction actually spikes using internet-native signal — attention liquidity, memetic traction, market-priced narrative.
04Selective reinforcementChannel capital and support hard into the shoots with exponential uptake. Prune on a cycle; the pruning itself makes content.
05Nutrient redistributionReturns feed shared infrastructure and the living network, not only the fund — so every strand strengthens the next probe.
The nutrient field

Internet cultural markets as the signal layer

Traditional venture signal is school brand — heavy, high-gatekeeping, and fundamentally unscalable. Internet-native signal is memetic traction and attention liquidity: globally reachable, continuously priced, and available long before a warm introduction is.

The soil is not limited to Silicon Valley gardens. The nutrient gradient is the internet itself — and the people earliest to reading a paradigm shift capture most of the alpha, if they are watching the right instruments.

The structure

A self-building XPrize

XPrize is a top-down fruiting body: declare the grand challenge and hope something grows into it. Half a billion in prizes has drawn only tens of thousands of participants, because the bottom-up funnel was never built.

Invert it. Announce the mission at the top, and let the projects form organically underneath, each competing for the resource stream — with an evaluation metric that asks not only what capital returns, but how much does this accelerate the timeline of the mission.

The split

For profit & for prophet

Ideas sort into two columns: cash-flow-driven ventures, and ventures aimed at the future of meaning, belonging, community and currency. Persist separates into arms accordingly — which also creates category leaders and allows higher overarching ownership across multi-company initiatives.

MediaAIRoboticsCrypto

Four primitives  ·  we believe these are core to the rewiring of society, and that all large returns come from a foothold in the disruption they create.

The frame

Spear and ritual

Two triads keep the portfolio legible. The spear is CAM — Community, Altruism, Media: how a venture earns the right to attention. The ritual is ARC — AI, Robotics, Crypto: the technical substrate it is built on.

The working intent is for every venture to be describable as one of each. Where it can’t be, that is usually a sign we don’t yet understand what we’re funding.

06  /  Bets

Meta-bets, and the honest probability each one needs.

The portfolio contains first-principles bets ranging from ten billion to a trillion dollars of potential value capture. The act of succeeding is, largely, continuing to recruit and continuing to drive narrative.

Rather than assert confidence, it is more useful to state the threshold: above what probability does each bet become positive expected value? These are low bars, and that is the entire point of the structure.

The betOutcome valuePositive EV above
Open Droids raises $50M at a $500M valuationOpen-source robotics  ·  majority-held
$500M
6.15%
Bump leads music-fi and becomes the Nasdaq of MusicThe designated proof case for the conversion ratio
$100B
0.0607%
Persist becomes the next AlphabetThe conglomerate outcome for the holding company itself
$3T
0.000667%
Open Droids establishes the robotics reserve currencyThe tail bet — deliberately included as a tail
$100T
0.00008%

Each threshold is Persist’s ownership share of the outcome value divided into the holding company’s own carrying value — the probability at which the bet breaks even. Ownership is majority in Open Droids and roughly a third in Bump. These are break-even probabilities, not forecasts: the claim is about the shape of the bet, not the odds of any single outcome. Read the bottom row as a tail included for completeness rather than a modelled scenario.

Focus

Open Droids

Open-source robotics, and the current all-in. The stated ambition is explicit: winning the robotics narrative — specifically deployment — is the one outcome that would make Persist the most valuable holding company in the world.

Strategy  ·  own the deployment story: developer app store, published frameworks, a package the robotics community actually adopts, early SEO in a market that hasn’t been claimed.

Proof case

Bump.fm

The designated proof case, and not yet proven. The claim the whole model rests on is that roughly $30K of cash converts into eight-figure validated equity — and that becomes a fact rather than a projection the moment this entity closes a sizable seed. It is the closest to doing so.

Will prove  ·  the conversion ratio the entire model rests on. Until the round closes, treat it as a projection.

Platform bets

Vible · Wemotions · Synergy

A bloom-scroll feed, a genuinely social social network, and automated co-founder discovery. Read alongside the thesis, each is a candidate platform for exactly the kind of participatory feedback that argument says will matter most.

Shared challenge  ·  the cold start. Rather than solving it separately per product, we solve team formation and launch mechanics once — as we did with the original funnel.

Placements — the credential pairing, live

Mend

An AI couples therapist, matched with a co-founder from a major operator network.

Chickstarter

Short-video-native crowdfunding, matched with a co-founder carrying celebrity and female-leadership relationships.

Chainreach

A creator-marketing platform, matched with a co-founder with deep creator-economy access.

Early evidence for the 2.0 correction: credentialed talent of this calibre is visibly interested in the model — two selections out of fifty-plus conversations so far — and the bet is that their arrival is what turns a good product into a company with a real valuation.

07  /  Trajectory

Five years, stated as a floor and a ceiling.

Two numbers per year, not one. The lower bound is the line below which we would consider ourselves to have underperformed. The upper bound is what we believe is reachable if the key moves land — and it is deliberately not the base case.

Net asset value, 2026–2030

Log scale  ·  USD  ·  range per year

Upside case Stated floor

Plotted on a log scale because the range spans three orders of magnitude; a linear axis would flatten the near years into the baseline. The starting point is projected NAV to date. Both bounds are stated targets rather than commitments — the lower one is the level below which we would judge ourselves to have underperformed.

Net asset value, floor and upside, USD
YearStated floorUpside caseSpread
Today$115M projected——
2026$235M$3B12.8×
2027$550M$6B10.9×
2028$1B$15B15.0×
2029$1.5B$50B33.3×
2030$2.5B$100B40.0×
Unlock 01

One sizable seed

The first fund proves itself the moment a single sub-entity raises a real seed round — demonstrating that a small cash outlay converts into eight-figure validated equity. Everything above the floor line depends on this happening more than once.

Unlock 02

A head operator

The current model deserves more focus than one person can give it while also driving the highest-conviction venture. The search is for a head operator with a prior exit, real connections, and the same conviction — and there is meaningful equity attached to finding them.

Unlock 03

Public credibility

An announced round does three things at once: it draws inbound senior leadership, it makes the ecosystem legible to partners, and it converts the story into the recruiting asset that the head of the snake actually needs.

The probability of total failure is near zero. At any time this can stop exploring and become a profitable company.

On the downside case

Why the downside is structurally bounded

The scale of a venture fellowship is not capped by the addressable market of any single company — it becomes what any large company eventually becomes, a conglomerate of startups under a shared umbrella. Starting from that premise makes us more efficient at filling the many niches of the future than an entity that arrives at it by accident.

The real cost of the model is different and worth naming: attention is spread across teaching, nurturing and learning how to build the machine, rather than concentrated on one hyper-effective company. Whether that is the cost or the entire point depends on how much you believe the machine generalises.

08  /  Idea Map

How the ideas interlock.

Twenty-two core ideas, and the dependencies between them. Select any node to see what it rests on and what it enables — the argument is a network, not a list.

Select a node
Origin Model Thesis Machine Phase 3.0

In an age of magic, find the wizards.

Give them wands
Give them robes
Show them spell books

The playing field is being levelled. Nobody has experience on the frontier. As technology scaling potential compounds, the systems that route high-value, persistent people onto exponential paths will win — especially when those paths are the ones less travelled.

We seek to create venture 2.0 — the recalibration machine.