The Fellowship a16z Built That India Needs More Than They Do
In early 2026, Andreessen Horowitz did something unusual for a VC firm.
They ran an experimental eight week fellowship. Not for founders or engineers but for storytellers.
65 people. Selected from over 2,000 applications. Eight weeks of intensive training in narrative strategy, distribution mechanics and how to make technology culturally relevant.
By April, they had already launched Cohort 2. Another 65 fellows. They’re calling it “the Thiel Fellowship for the terminally online” But that’s not what caught my attention. What caught my attention was the gap it revealed. And once you see it, you can’t unsee it.
a16z built infrastructure for narrative scarcity. India has ten times the scarcity, none of the infrastructure, and every reason to move first.
What a16z Actually Built
The New Media Fellowship isn’t a marketing bootcamp. It’s infrastructure for a specific kind of scarcity.
Fellows aren’t people who make videos or write tweets. They’re the layer between “content creator” and “company storyteller” , people who can drop into a startup for two weeks, orchestrate a launch that captures mindshare, and leave behind the muscle memory for the team to keep doing it themselves.
They come from everywhere: MrBeast’s production team, HBO’s Silicon Valley writers’ room, marketing leads at @vercel and @OpenAI , growth operators from hypergrowth startups.
The output is immediate and measurable. Fellows get forward-deployed into portfolio companies. They run launch campaigns, build owned media channels, and often stay full-time or spin out their own ventures.
"Highly functional people who are also very online," as a16z's Brent Liang put it.
Fellows get forward-deployed into portfolio companies during the program. Some run launch campaigns. Others build owned media channels. Many end up staying full time. A few spin out to start their own. All of them become permanent nodes in a16z’s narrative network.
Real portfolio launches included @FAL ($125M at 1.3M views), @sola_ai ($17.5M at 319K views), Elise AI ($250M at 243K views). Not marketing. Mindshare.
Launch-as-a-Service (LaaS) framework the productized playbook fellows deploy inside companies. It works because a16z treats narrative capacity as a scarce, investable resource exactly like engineering talent or design chops.
It works because it treats narrative capacity as a scarce resource worth investing in, just like technical talent or design chops.
“Turning VC into a network effect business”, the strategic logic behind the fellowship. This isn’t charity. It’s infrastructure that compounds for decades.
This isn’t goodwill. It’s a deliberate play to turn venture capital itself into a network-effect business, with narrative infrastructure as the moat no other firm can copy overnight.
Portfolio companies like Abridge, Anduril, Applied Intuition, Databricks, Decagon, Deel, ElevenLabs, Flock Safety, Kalshi, Quora, Replit, World Labs is a slice of the a16z portfolio the fellowship serves. Every one of these companies gets a forward-deployed narrative operator when they need one.
And India doesn’t have anything like it.
What Is Price’s Law?
In 1963, a physicist named Derek J. de Solla Price was studying scientific publications, trying to understand why some researchers dominated their fields while others published and got zero attention.
He noticed something strange. The distribution of productivity wasn’t a bell curve as you’d expect. It wasn’t even close.
It followed a completely different mathematical pattern.
Price’s Law states that the square root of the number of people in a domain does 50% of the work.
Here’s what that looks like in practice:
In a company with 100 employees, 10 people produce half the output
In a field with 10,000 scientists, 100 produce half the meaningful research
On a team of 25, 5 people carry the entire operation
The formula is simple: √n = your high performers, where n is the total population.
Oh, and it wasn’t exclusive to research papers and this pattern shows up everywhere once you know what to look for.
Once you see it, you can’t unsee it.
Spotify has 11 million artists, but 50% of all streams come from only 3,300 of them.
Of the 30 million businesses in the United States, about 5,500 (the square root) generate half the total economic output. Amazon, Apple, Microsoft, and a few thousand others produce as much as the other 29,994,500 combined.
YouTube: very few channels account for the vast majority of both views and ad revenue.
River systems, sales teams, Wikipedia editors, wealth distribution — anywhere you look, the square root does half the work.
This isn’t coincidence, rigged systems, or unfair advantages (though those exist too). This is just how complex systems behave when skill, consistency, opportunity, and luck compound over time.
And if you’re trying to understand why Indian tech can’t capture cultural mindshare despite having more raw ingredients than almost anywhere else, this law might be the clearest lens you’ll find.
India’s Strange Paradox
India has stronger tailwinds than almost anywhere on earth right now.
The Union Budget 2026-27 allocated ₹250 crore for AVGC Content Creator Labs across 15,000 schools and 500 colleges, with IICT Mumbai as the national nodal agency. The explicit goal: 2 million AVGC professionals by 2030.
India supplies 15.2% of all global Web3 developers, the fastest-growing major hub on earth. AI × Web3 startups represent 33.2% of new Indian Web3 companies, the highest concentration globally.
@localhosthq closed a $2.5M angel round in February 2026 to scale founder labs in Bengaluru. The creator economy already influences $350–400 billion in consumer spending on track for $1 trillion+ by 2030.
But here’s what happens when Indian tech actually launches:
- A promising protocol launches → generic PR deck.
- A hackathon runs → recycled Twitter threads. - A well-funded startup raises capital → silence, or jargon-heavy posts no one reads.
We have developers. We have capital. We have government backing. We have events.
We don’t have storytellers.
And it’s not just Web3. Indian AI startups can’t explain their models. SaaS companies with global traction get ignored at home. Climate-tech founders build world-class solutions with zero cultural resonance.
The bottleneck isn’t innovation. It’s narrative infrastructure.
Competence vs. Incompetence
Price’s Law violates our egalitarian instincts. We want to believe everyone contributes equally, that effort equals outcome, that hard work is the great equalizer. But reality is far from it.
Real storytelling competence is rarer than we pretend it is.
I’m not saying Indian creators are lazy or untalented. But real narrative competence, the kind that moves markets, sparks launches, and makes a hard technical concept suddenly click for millions is the result of skill, consistency, opportunity, and some luck, all compounding in the same direction.
Most creators have one or two of those ingredients. The square root has all four.
India has roughly 2-2.5 million monetized creators. Only 8-10% of them monetize effectively.
The square root of those who actually matter for tech narrative? Scattered across Telegram groups, Substack newsletters, and WhatsApp communities with no structure to channel them.
They exist. We just have no way to find them, no way to train them, no way to put them in the room with the founders who need them most.
And here’s the uncomfortable part: even if you tripled the number of Indian creators tomorrow, Price’s Law says the output gap wouldn’t close linearly. You wouldn’t get 3x the narrative impact. You’d get √3x, about 1.7x.
The only way to move the needle is to find and leverage the square root directly.
That’s exactly what a16z figured out.
16z’s fellowship model isn’t about training everyone. It’s about finding the square root and giving them leverage.
Here’s the structure for an Indian version:
Eight weeks. 30–40 fellows. Hybrid format across Bangalore and Mumbai.
Weeks 1–2: Frameworks. Storytelling arcs for technical products. Distribution mechanics across X, LinkedIn, Telegram, newsletters, short-form video. How to separate what’s actually interesting about a launch from what founders think is interesting.
Weeks 3–4: Real embeds. Each fellow paired with a participating Web3, AI, or deep-tech startup. Real launches. Real timelines. Real pressure. The fellow runs narrative strategy and trains the internal team.
Weeks 5–6: Content sprints. Ship real assets. Video series. Multi-part essays. Telegram campaigns. Podcast verticals. No mockups. No drafts.
Weeks 7–8: Public showcase at a major national tech or creator event. Best projects get featured. Best fellows get job offers. Best ideas get funded.
Biweekly dinners alternating between Bangalore and Mumbai. Workshops on virality mechanics, Telegram-native community design, vernacular scale, AI-augmented storytelling. A tight group chat that becomes the “terminally online” network for Indian tech.
Why India Does This Better Than a16z
This isn’t a copy-paste of the US model. India has structural advantages no US fellowship can replicate.
Multi-chain and ecosystem-agnostic. Not tied to one firm’s portfolio. Fellows train across Ethereum, Solana, Polygon, Aptos, Cosmos.
Vernacular + AI-native superpowers. A fellow who ships Hindi, Tamil, or Telugu AI-generated explainers reaches Tier-2/3 India in ways English-only operators never will.
Telegram-native distribution. Where real conversations happen in Indian crypto, AI, and creator communities. 61% of new crypto traders in India are Gen Z. They don’t check Twitter. They check Telegram.
Culture bridges that matter. IPL → prediction markets. UPI → stablecoins. Bollywood → creator economy dynamics. Reference points that make hard tech click in ways Silicon Valley analogies never will.
The Paradox (And Why It’s Not Optional)
In an ideal world, we’d all know the best moves to make at any given time. But that’s not how the world works.
You can’t know which 10 fellows out of 30 will become India’s narrative square root without running the program. You can’t know which embeds will produce breakout launches without shipping them. You can’t know which frameworks translate to Indian context without testing them.
It’s a game. We all have to play.
The early game is exploration. You’re planting seeds. Price’s Law hasn’t kicked in yet because you don’t have enough data. That’s Cohort 1.
The middle game is identification. You’ve shipped 30-40 fellows through the program. You’ve run 10-15 live embeds. You’ve produced real campaigns with measurable outcomes. Patterns emerge. This is what works. That’s Cohort 2.
The late game is exploitation. You double down on the winners. You build the alumni network. You turn successful fellows into mentors. You let Price’s Law work for you instead of against you.
Most ecosystems never leave the early game. They keep running disconnected bootcamps, scattered creator grants, and one-off media experiments, mistaking activity for progress.
Others skip straight to the late game. They announce a massive national program before they have any data. They commit to a strategy with a sample size of zero.
The paradox is you need both. Explore the noise until you have signal. Then exploit that signal hard.
So let’s run the pilot. Ship the first cohort. Publish the “bad” campaigns alongside the winners. Take the pointless meetings until the right ones surface.
Because you really don’t know what’s useless until after the fact.
Compounding happens in the margins, in the weird micro-connections between the stuff that didn’t work and the stuff that did.
India has the government funding. The private momentum. The talent. The events. The creator economy tailwinds.
What we don’t have is someone willing to press “go.”
Competence is rarer than we pretend it is.
There are fewer of you.
Let’s find them.
Thanks for reading. If this resonated please tag someone building in Indian tech who should see this.