AI, Healthcare, VC..., 10 Charts Capturing Changes in the World

84% of humans have never used AI, and only 0.3% pay for it.

Author: Rex Woodbury

Translation: Deep潮 TechFlow

Deep潮 Guide: Rex Woodbury of Daybreak Ventures outlines key trends for early 2026 through 10 charts.

Key Findings:

(1) Newspaper stocks plummeted five years before earnings declined; SaaS stocks are now repeating this pattern;

(2) 84% of people have never used AI, with only 0.3% paying for it;

(3) AI application usage time has surged, competing with Netflix and TikTok for time;

(4) Healthcare accounts for 15% of US employment, driving nearly 100% of job growth;

(5) The secondary market is growing alongside IPOs and M&A;

(6) Gen Z’s “financial nihilism”: first-time homebuyers average age is 40, more like gambling;

(7) Retatrutide (Lilly’s new drug) could be a trillion-dollar peptide;

(8) Gaming market worth $200 billion, Roblox user engagement surpasses Steam, PlayStation, and Fortnite;

(9) 50% of Anthropic agent calls are in software engineering;

(10) Citrini research report triggers market panic selling — we live in a meme economy.

Full text below:

It’s that time again: time for another “10 Charts” series.

I try to do this quarterly, but we’re overdue: the last one was in October. This is our 11th edition (!). You know the rules: I’m a visual learner, and charts help me process information. Charts are also an effective way to show how the world is changing.

We’ll cover 10 charts on a wide range of topics:

1. Newspaper stocks vs. earnings
2. Still the first game
3. AI application usage time
4. Healthcare driving employment
5. Reshaping VC and employee returns in the secondary market
6. Gen Z: the last generation in the alphabet
7. Peptides and Reta
8. Gaming landscape
9. Calling for more agent calls
10. Citrini sell-off

Without further ado…

Newspaper Stocks vs. Earnings

This chart shows the comparison between newspaper stocks and earnings. You can see that stocks crashed about five years before earnings declined, indicating the market saw the writing on the wall before it hit the income statement.

Source: Twitter; thanks to Emily Man for sending

Of course, there are some timing discrepancies; the decline in forward earnings coincided with the Great Recession. But in terms of direction, it seems accurate—the market anticipated the internet disrupting newspapers. We’re seeing this again now, with SaaS stocks crashing last month before AI disruption.

As we wrote last week, SaaSpocalypse (the SaaS apocalypse) will take time to unfold. I attended a panel discussion this week where a member said “Campbell’s Soup won’t vibe-code their own CRM,” a clever way to put it. But the market is pricing in a final reality: software profits are compressing, and the “new normal” is 70% gross margin instead of 90%.

On the topic of AI taking a long time to fully develop…

Still the First Game

This is a cool visualization of where we are in the AI adoption cycle. Each dot represents 3.2 million people. There are 2,500 dots, totaling 8.1 billion people.

• Gray = 6.8 billion who have never used AI
• Green = 1.3 billion free users
• Yellow = 15-35 million paid users
• Red = 2-5 million coding users

Source: Noah Epstein on Twitter

Approximately 84% of the world has never used AI, and only 0.3% (!) pay for AI products. This is one of the best visualizations I’ve seen showing how early we still are.

AI Application Usage Time

In the previous “10 Charts” edition, we looked at ChatGPT’s “smile curve”:

From that article:

This chart commits what I call the “Y-axis crime,” meaning the Y-axis is misleadingly not starting at zero. But in this case, the Y-axis crime actually hurts ChatGPT! When you realize the worst group is approaching a horizontal asymptote in their 40s (then “smiling”), the curve looks even better.

These types of curves are usually reserved for markets with network effects or social products, meaning as more users join, they improve (e.g., Uber with more drivers/riders, or Instagram with more friends). For a solo product without social features, such retention is impressive.

Besides improving retention, AI apps are also seeing increased engagement. Here’s a visual showing this trend:

Overall, this is a very impressive growth in usage.

In 2017, Netflix’s Reed Hastings famously said the biggest competitor to Netflix is sleep. Netflix’s business is about absorbing more viewing time (more hours watched = better retention + willingness to pay), so sleep directly conflicts with the business model.

Now we see media consumption leveling off at about 12.75 hours per day:

The increase in AI usage must come at the expense of time spent elsewhere. Maybe Claude’s biggest competitor is sleep? I also imagine Netflix, YouTube, TikTok are watching the above AI usage charts with caution. Half an hour on Gemini is half an hour not watching short videos. AI tools are clearly more than just Google replacements; they’re also social + content products. Once generative media really takes off, we’ll see huge pressure on engagement metrics of existing giants.

Healthcare Driving Employment

Healthcare is the largest employment category in the US, accounting for about 15%. It’s the engine behind nearly all job growth. Check out this chart:

Overall, healthcare will drive about 40% of new jobs over the next decade. Meanwhile, the fastest-growing single job in the US is “home health aide,” driven by rapid aging of the population (10,000 Americans turn 65 every day).

Healthcare benefits from several tailwinds:

1. LLMs are well-suited for healthcare management, a trillion-dollar market because healthcare relies on language.
2. Consumers are increasingly willing to measure, personalize, and spend on their health.
3. Telemedicine is expanding access, aided by new regulations post-pandemic that broaden coverage.
4. Our population is aging and becoming sicker—“silver tsunami,” and so on.

Many healthcare jobs are also “AI-protected,” which I interpret as meaning we’ll see more young people entering healthcare.

Reshaping VC and Employee Returns in the Secondary Market

This is an underestimated shift in venture capital. The secondary market now rivals IPOs and M&A exits:

Source: Tomasz Tunguz on Twitter

This changes the game for early-stage funds like Daybreak and startup employees. Liquidity timelines are compressed. I wouldn’t be surprised if we see funds returning multiples of their invested capital through secondary sales during growth rounds. It’s not new—Roger Ehrenberg of IA Ventures has publicly discussed selling about 2.5 million shares of The Trade Desk (out of roughly 6.6 million) in secondaries to return capital to LPs—but it’s becoming more common.

Employees no longer have to wait over 10 years for liquidity. Clay and ElevenLabs each completed two acquisitions in the past 12 months, and Anthropic is currently in a $6 billion (!!!) acquisition process. The latter will undoubtedly shake up the SF real estate market.

Gen Z: the Last Generation in the Alphabet

Kalshi reports that Super Bowl betting volume exceeded $1 billion, a 2,700% increase year-over-year (!). Here’s a chart of predicted market trading volume for the Super Bowl, showing a 1,205% increase over six months:

These markets are new and controversial. White House press secretary Karoline Leavitt abruptly ended a briefing in early January, sparking insider trading concerns:

During the Super Bowl, my partner placed some bets on Cardi B’s appearance on stage with Bad Bunny. He lost that money, and Kalshi said performers must sing to count as “performance.” This led to at least one complaint to the CFTC. It’s the wild west!

But despite the controversy, I believe prediction markets will stick around. Last fall, we wrote about the rise of prediction markets in “Speculation Nation.” That article focused on enabling technologies colliding with Gen Z behaviors, including the rise of FAFOnomics (FAFO = Fuck Around and Find Out).

My friend Jackson Denka wrote an interesting piece this week titled “Financial Nihilism or: How I Learned to Stop Worrying and Love the Market.” He calls Gen Z “the last generation in the alphabet,” which struck me. Some stats he cites:

• Unemployment among US college grads in 2025 is 9.3%, higher than during the Great Financial Crisis
• The top 1% of households hold nearly 30% of the country’s wealth
• The average age of first-time homebuyers is now 40

No wonder we’re becoming a speculative economy. If upward economic mobility is Sisyphusian, why not bet everything for a shot at wealth? Note: this isn’t a good thing, but I see it as one of the defining undercurrents for the next generation.

Peptides and Reta

Amid the AI buzz, it’s easy to overlook other seismic shifts. I’ve been spending a lot of time on one area: peptides, which are starting to gain mainstream attention.

Peptides are signaling molecules made of amino acid chains in your body. The most famous are Ozempic and Wegovy, with the peptide semaglutide. The peptide market is booming because consumers are genuinely interested and willing to pay. My friend Khushi said it well in this tweet:

Our first 2026 investment is a peptide company, System Labs, which launched last week. Making peptides accessible, trustworthy, and safe for everyday consumers in the US has huge potential.

Here’s the market’s expected growth:

The most undervalued drug right now is Retatrutide, or Reta. Lilly’s Reta is a triple agonist, while Ozempic is a single agonist; it’s a fancy way of saying Reta targets three receptors: GLP-1, GIP, and Glucagon. This means the drug enhances satiety, improves insulin sensitivity, and boosts metabolism (burns fat). Ozempic only targets GLP-1, mainly reducing appetite.

Reta is a potential trillion-dollar drug. Here’s a chart of Reta’s weight loss effects:

Source: CTCD

Looking forward to hearing more about Reta soon.

Gaming Landscape

Matthew Ball released a lengthy report on the state of gaming last week. Here are some highlights:

Gaming remains the largest media category, with annual spending of $200 billion, surpassing movies, TV, and music combined. After a brief dip post-pandemic, growth has resumed:

Mobile is driving most of the growth:

Despite market growth, VC funding has declined significantly since the pandemic peak:

AI will reshape gaming, though it’s not fully there yet (we’re still mostly in the text phase). Soon, game generation rather than rendering will become standard, opening new possibilities for storytelling and world-building.

Returning to our earlier point about AI competing with sleep and Netflix: we may also see AI applications encroach on gaming time. This tweet resonated with me:

The most impressive company in gaming remains Roblox, which has driven a large part of the growth:

Roblox now has 150 million daily active users, with rapid growth among users over 13:

Roblox’s average engagement now surpasses the combined total of Steam, PlayStation, and Fortnite.

Calling for More Agent Calls

Here’s an interesting chart showing Anthropic agent calls across industries:

Clearly, opportunities outside software engineering are huge. Or as Garry Tan put it:

As for what happens to these jobs: I think the workforce impact will be slower. We’ve written extensively about Jevons’ paradox here. Here’s a good visual from Coatue about ATM jobs:

People thought ATMs would destroy bank tellers. Instead, from 1970 to 1988, the number of tellers increased 81%, enjoying four decades of steady growth.

Citrini Sell-off

Another week, another viral blog post. This week, it’s a Citrini Substack article that triggered market sell-off:

What’s crazy to me is that a casual article about the future, with no facts or data, can cause such widespread market panic. To me, it’s a sign that (1) markets are overheated and looking for reasons to correct, and (2) we’re officially in a meme economy.

I find the article naive. Fintech and market companies are harder to disrupt than Citrini suggests. Tony Xu of DoorDash had a good response.

The real lesson from the Citrini sell-off: nobody knows what will happen.