Time to Humanity

We used to call it time to fun.

At Epic, I ran creator partnerships for Fortnite. $200 million in portfolio revenue, 400 million monthly players, Travis Scott performing inside the game. We tracked how long it took a new player to hit their first moment of genuine enjoyment. The tutorial didn't count. The first match didn't count. What counted was the moment their brain lit up and decided to come back. Everything between the loading screen and that moment was friction. Our job was to kill the friction.

The metric never got a Wikipedia entry. It lived in production meetings and Slack channels, the kind of vocabulary that circulates among people building live-service games but never gets published. Jaime Griesemer had a version at Bungie in 2002: thirty seconds of fun that repeat forever. Celia Hodent, who built Epic's UX practice, turned the first-time player experience into a science. The whole philosophy boiled down to one question: how fast can we get you to the thing that matters?

Epic eventually added bots to new-player lobbies. More than half the enemies in your first matches weren't real. You didn't know it. You got kills, you got dopamine, you came back. Seventy-eight million unique players in a single month. Billions in revenue. Time to fun wasn't a vanity metric. It predicted everything.

I left gaming. But the measurement stuck. The idea that you could quantify the distance between someone and the thing that matters, and then shorten it.

Every industry discovers its version. SaaS calls it time to value. Facebook's was seven friends in ten days. The pattern is always the same: measure the distance between the user and the thing that matters most, then eliminate everything in between.

Here's what I can't let go of. When production costs collapse to zero, when anyone can generate anything, when half the internet isn't even human anymore. What's the thing that matters most?

Proof that a human was involved.

The man the internet trusts most for geopolitical predictions isn't on Bloomberg. He's a high school teacher in Beijing. No university appointment. No monetization. His lectures went viral after his calls started coming true, and the audience didn't care about credentials. They watched a man think in real time in a room full of students. The classroom was the verification.

That instinct, the immediate trust calibration based on context rather than credentials, is showing up everywhere. Most people just haven't named it yet.

Bots now account for 51% of all global web traffic. Seventy-four percent of newly published web pages contain AI-generated content. Reddit's co-founder said it plainly: “the dead internet theory is real.” The background noise of the internet is no longer human.

Against that backdrop, the fakes are getting specific. Aitana Lopez has 149,000 Instagram followers, brand deals with Victoria's Secret and Amazon, and pulls $30,000 a month. She doesn't exist. She was built by a Barcelona agency tired of dealing with real models. “Slop” was Merriam-Webster's word of the year in 2025. The cultural immune response landed on contempt.

The tools cut both ways. ChatGPT costs twenty dollars a month. A Five Guys burger. CarEdge built an AI car negotiator for forty dollars a month that saves buyers over a thousand dollars per deal. Two-thirds of Americans who've used generative AI have used it for financial advice. More than a third consult ChatGPT before meeting their real-life financial advisor. Your track record used to be your moat. Now it's your attack surface. Twenty dollars a month buys anyone a personal analyst that majored in calling bullshit.

So is any of this measurable? Or just a feeling?

OnlyFans did $7.2 billion in gross revenue in 2024. Every AI companion app on earth, combined, did $120 million. Same market. Same buyer motivation. Parasocial connection, intimacy, the feeling of being known. One has verified humans. The other has algorithms.

That's not a rounding error. This is what the market values.

Time to Humanity: the speed at which someone can verify that a real person, with real judgment and real stake, is behind what they're seeing.

It used to be free. Everyone assumed a human made everything. The cost of proving your humanity was zero, so it contributed nothing to competitive advantage. Then generative AI collapsed production costs toward zero for text, images, code, video. The supply of “good enough” became infinite. And the thing that was previously free, proof that a human actually made this, suddenly became the scarcest input in the equation.

The shift was not gradual either. It underwent a phase transition. Same molecule. Completely different behavior.

The verification premium

The obvious objection: if people can't tell the difference, why would any of this matter?

The New York Times ran a blind test in March 2026. Eighty-six thousand participants. Fifty-four percent preferred the AI-generated writing. For factual reporting, AI preference hit sixty-two percent. Detection accuracy across three countries has declined to chance levels. A coin flip.

Which sounds like the thesis is dead until you realize what happens when you add a label.

Present identical content. Tell one group a human made it. Tell the other group AI made it. Preference for the human-labeled version jumps thirty percent. The Journal of Business Research found that content believed to be AI-written triggers what they called “moral disgust,” reducing purchase intent. People devalue AI-labeled art even when they've already said they can't distinguish it.

People prefer AI in blind tests. And then they'll pay thirty percent more the moment you tell them a human made it. Same content. Same quality. Different label. The label is doing all the work.

The rebuttal writes itself: novelty response. Early audiences rejected CGI too, and now nobody notices. But organic food premiums have held for twenty years despite most consumers admitting they can't taste the difference. The premium persists because it's structural. Insurance mandates, regulatory disclosure requirements, and platform demonetization policies are making human verification institutional. The 30% is the early reading of a curve that's steepening.

TTH is a verification premium. The scarce thing isn't the human. It's the proof of the human.

The repricing is already visible. Twenty-nine companies. Two outcomes.

Methodology: 29 companies selected across sectors to illustrate the TTH repricing pattern. X-axis positions are qualitative assessments by the author. Y-axis values are publicly reported P/E, market cap, or revenue changes from January 2022 to present. This is illustrative, not a controlled study. The pattern is directional. Multiples moved for compound reasons — TTH exposure explains the divergence within industries, not the absolute levels.

These multiples moved for compound reasons. Luxury demand cycles. Post-COVID spending. Platform algorithm changes. Debt loads. But compound reasons don't explain divergence within the same industry. The New York Times and BuzzFeed are both digital media companies. Both faced the same AI disruption, the same ad market, the same platform dynamics. One invested in editorial identity and journalist bylines. The other pivoted to AI-generated content. One expanded 35%. The other lost 92%. Same industry. Same disruption. Different TTH.

The Accenture row is the nuance. Consulting is high-human-involvement, but the market sees billable hours as AI-replaceable. Accenture compressed 48%. Hermès expanded 18%. Hermès makes things that can't be automated. Accenture sells hours that can. The market is drawing that line in real time.

The strongest counter: infrastructure like C2PA will make verification cheap and ubiquitous, collapsing the premium before it consolidates. But credit ratings didn't get cheaper after EDGAR made financial data free. They got more valuable, because the interpretation layer matters more when raw data is abundant. When everyone can stamp content with provenance metadata, the question shifts from “is this verified?” to “who do I trust to tell me what the verification means?” That's where the oligopoly forms.

Who captures the value

Every time something free becomes scarce, a verification business emerges. Moody's rates bonds. Three firms control 95% of all credit ratings. Operating margins above 60%. The bond market is $145 trillion. They capture a fraction of a percent and control pricing for the entire market. The pattern is structural: verification businesses tend toward oligopoly, operate at extraordinary margins, and outlive the commodities they certify.

Nobody has built the Moody's of humanness. But the EU AI Act mandates disclosure by August 2026. C2PA has 6,000 member organizations. Eighty-plus proof-of-personhood startups are competing. The market will consolidate to two or three. It always does.

And there's an accelerant. Major insurance carriers are seeking to exclude AI liabilities from standard policies. AIG, Berkley, Hamilton. Verisk is rolling out standardized exclusion forms effective January 2026. If human-authored output is insurable and AI-authored output increasingly isn't, verification stops being cultural preference and becomes economic infrastructure.

Three implications. One for your money. One for your time. One for your head.

For your money, look where TTH is newly scarce and the market hasn't caught up. Hermès has always traded on human craft. That's priced in. The opportunity is in companies where TTH just became their most valuable asset and nobody's noticed.

Streaming platforms are in an open auction for human personality. xQc signed a $100 million Kick exclusivity contract. Kick offers 95/5 revenue splits versus Twitch's 50/50. Kai Cenat's twelve-hour subathon generated $1.2 million. The asset being bid on is a verified human presence that can't be replicated. The platforms know what TTH is worth. They're paying for it in nine figures.

Big Tech is pricing individual humans even more explicitly. $40 billion in acqui-hire deals in 2024–2025. Microsoft paid $650 million for Inflection's seventy people. Google paid $2.7 billion for Character.AI, primarily for two founders. Meanwhile WPP, the world's largest ad holding company, trades below 5x. Talent-dense boutique agencies sell at 8–12x. Same industry. The market is separating the humans from the wrapper.

I wrote about a specific version of this dislocation in The Great Data Inversion. Time to Humanity is the general theory underneath it.

For your time, invest in activities that lower your personal TTH. Writing with real detail attached to your name. Publishing where people can verify your judgment through direct experience. Building relationships where the proof accumulates over years, not followers.

I built an AI to help me write this. Her name is Paris. She manages my inbox, tracks my deals, runs my research, reminds me to journal. I coded the first version myself in two weeks and I've rebuilt her three times since. She's the best analyst I've ever worked with and she cannot write this paragraph.

The play isn't rejecting AI. I use it more than almost anyone I know. The play is knowing which layer is yours. Paris can surface every data point in this essay. She can't decide which ones matter, or what it feels like to measure something you used to take for granted. The proof of humanity isn't refusing the tool. It's knowing what the tool can't replace about you specifically.

This site is the proof of work. Every piece is a bet that the proof accumulates. Reduce the friction between you and proof that you're real.

For your consciousness, the carousel I wrote about in the Consciousness Ponzi is a TTH crisis. Crypto bro, trades evangelist, born-again, looksmaxxing. Every identity on the rotation is easy to perform and hard to verify. High time to humanity. Quick to adopt, quick to abandon.

Staying with one thing long enough that the proof accumulates is the hardest move in the playbook. Accumulating authenticity instead of performing it. I'm three pieces into a site nobody asked for, writing about things I can't stop thinking about, and I still don't know if it's working. But I know the metric. And I know the friction is lower than it was yesterday.

I think about that teacher in Beijing. No credentials. No monetization. Just a man thinking out loud in a room until the proof accumulated. We measured time to fun because it was the only metric that predicted whether someone would stay. Time to humanity works the same way.

The market already voted. Sixty to one.