Ben & Marc: Why Everything Is About to Get 10x Bigger

The single most common mistake in investing is sizing markets based on what exists today. Every major technology breakthrough proves the same lesson: supply creates its own demand, and the real market is orders of magnitude larger than the one you can measure.

The Conversation

a16z cofounders Marc Andreessen and Ben Horowitz sit down with general partner Erik Torenberg and Not Boring’s Packy McCormick, who had just published an in-depth profile of the firm. The conversation ranges from how the media ecosystem has fundamentally shifted, to why a16z raised a $15 billion fund, to what makes Zoomer founders different from every previous generation. The tone is unusually candid for a firm that’s often accused of being “promotional.” Marc and Ben are reflective, occasionally self-deprecating, and clearly energized by what they see coming.

Supply Creates Demand: The 10x Cloud Multiplier

Marc keeps returning to a single framework: supply-driven markets. The classic investor instinct is to size the addressable market by looking at what currently exists, then estimate how much of that a new company might capture. But this fails catastrophically when there’s a fundamental supply-side breakthrough.

The examples stack up quickly. The market for Uber wasn’t the taxi market. The market for cloud software wasn’t on-prem licenses. The market for GPUs wasn’t gamers. In each case, the technology change unlocked demand that simply didn’t exist before, and the actual market turned out to be 10x, 100x, or 1000x larger.

Ben illustrates this with Databricks. When he was coaching CEO Ali Ghodsi, Ali was trying to convince a recruit that the company would be worth $10 billion. Ben’s reaction: why are you lowballing? The cloud data market was clearly going to be orders of magnitude larger than anyone was modeling. The limiting factor wasn’t demand; it was that nobody could see the demand because the supply hadn’t materialized yet.

“Nobody ever asked for a Macintosh. Nobody ever asked for an iPhone. These things had to be designed and built and provided on the supply side before the demand materialized.”

Marc frames this as the dominant investing trend for the next 30 years: significant technology changes on the supply side will keep unlocking much larger markets than traditional analysis can predict.

The Media Barbell: From Cat Videos to Three-Hour Podcasts

Marc applies the supply-driven framework directly to media. He walks through the moral panics of successive eras: television (“Americans watch six hours of TV a day, the vast wasteland”), then the internet (“people just want cat videos and short-form TikToks”). Each time, critics assumed consumers were fundamentally shallow.

But long-form podcasting shattered that assumption. People watch three-hour podcasts all the way to the end. The analytics prove it. The problem was never short attention spans; it was lack of supply. There’s enormous latent demand for smart, high-quality content across every domain, and Substack is the platform positioned to unlock it.

Marc argues Substack could be “a thousand times the size of the existing media industry” because the old centralized media structure was a constraint, not a reflection of actual demand. Writers felt trapped by media companies that undervalued them. When the supply constraint is removed, the market explodes.

This is precisely why a16z invested in Substack, using the same lens as their Databricks thesis: orders of magnitude larger than anything we’ve seen so far.

”We Reinvented the Computer”

When asked what a $15 billion fund signals about the future, Ben is blunt: “$15 billion to start.” The reasoning is simple. AI is a reinvention of the computer itself, and the new computer is far better than the one we’ve been building on for 50 years.

There isn’t a single problem they can think of at the firm that AI won’t be able to solve. Cancer, transportation, fraud. Name a problem in the headlines and the response is “we can solve that.”

Ben also expects the number of entrepreneurs to multiply because AI dramatically lowers the barrier from idea to product. The ease of building has fundamentally changed, which feeds back into the supply-driven thesis: more builders, more supply-side breakthroughs, more demand unlocked.

The Confidence Machine

The conversation’s most revealing section is about what a16z actually does for founders. Ben reframes the entire firm as a confidence engine.

Inventors are geniuses at building technology, which typically means they’ve spent 10-20 years in a lab. They’re fully capable of understanding the real world; they just haven’t encountered it yet. And the real world, as Marc notes, “is really, really big and really, really messy.” Eight billion people with opinions, many of them hostile to new ideas.

The recurring failure mode: inventors assume the breakthrough product will obviously sweep the market. If it doesn’t, they conclude the product isn’t good enough. But the actual problem is everything around the product: go-to-market, policy, hiring, recruiting, dealing with people who want to destroy what you’re building, or worse, just ignore you.

Ben describes a vicious confidence cycle that traps founders. They get bad advice from advisors who’ve never built anything. They can’t reach the important CEO, can’t recruit the top engineer, can’t navigate government. Each failure erodes confidence, which slows decision-making, which creates more failure.

The entire firm is designed to reverse this spiral. The brand, the connections, the policy team, the expertise, all exist so founders can borrow a16z’s power at the most critical moments in their development. As Marc puts it:

“The purpose of building the dominant venture brand was precisely to be able to have the companies borrow that at the most critical points in their development so that the companies can use our force in the world as a slingshot to build their own force.”

This also explains why a16z is so public-facing, so “promotional.” It’s not ego. It’s the battery that founders plug into.

Dream Builders, Not Dream Killers

a16z requires every new hire to sign a written culture document and sit through an hour with Ben understanding it. The core principle: they are dream builders, not dream killers. They are not the “analytical smarty pants who makes ourselves look smart by making somebody else look stupid.”

Anyone pushing the world forward, whether or not a16z agrees with their method or invests in them, gets support, not criticism. They will never “attack the future.” This extends to strict behavioral norms: never talk negatively about a technology, a founder, or a company in person.

Ben is clear about why this matters for reputation. One act of obnoxiousness or dishonesty causes more damage than five or ten good deeds. You cannot tolerate that behavior, which is why cultural enforcement is non-negotiable.

Reputation as the Compounding Asset

When asked what single thing a16z is compounding, Ben answers immediately: reputation.

He and Marc discussed this from day one. The evidence of compounding is concrete. Fund I was $300 million, took six months to raise, required countless meetings, and some LPs “didn’t treat us very well.” The $15 billion fund was raised after Marc did one AMA and Ben did one AMA. “I don’t know that I took another meeting on this fundraise. It was done on reputation.”

Reputation transfers directly to portfolio companies. When a startup takes a16z investment, they can use that reputation to break through with customers, recruits, downstream investors, and regulators. This is the entire point: reputation compounds for the firm and then transfers as leverage to founders.

Staying Small Inside a Big Firm

How does a16z avoid becoming the big company it warns founders against? Organizational design. The crypto group, infra group, apps group, American Dynamism group each operate like small companies with very specific support from the brand and fundraising infrastructure. Integration points between groups are simple. Each group is highly autonomous.

Ben says they borrowed heavily from the original Hewlett-Packard model, pre-computer-business: a series of companies inside a company. GPs can only lead a group after performing inside the firm for a long time. You can’t get that job from the outside. They’ve largely stopped hiring senior GPs externally and instead grow people from earlier career stages into the a16z culture.

The Mythical Man-Month Is Dead

Ben offers one of the sharpest observations about how AI changes venture dynamics. The mythical man-month, Brooks’ Law, was an iron rule for 50 years: adding engineers to a late software project makes it later. You couldn’t throw money at a technology lead.

With AI, you can. Elon Musk threw money at the foundation model problem and caught up “very, very fast.” In the entire previous history of the software industry, that could never have happened. If you’d asked an AI a year ago whether Elon had any prayer of catching OpenAI or Anthropic, the answer would have been no.

Ben’s deeper point: this means your highly trained pattern-matching neural net (as an investor) is increasingly dangerous. History is a bad guide because things change faster than your training data. The intangibles, the art of reading people, of making long-dated bets with uncertain outcomes, are becoming more important, not less.

Christopher Columbus and the Art of Venture

Marc connects venture capital to a pattern stretching back 500 years: a person with a dream, deep domain knowledge, and asymmetric payoff potential. Banks can’t underwrite a 50% chance of 10x and 50% chance of zero. Big companies won’t touch it. But when the dream works, especially across a portfolio, the expected value is very high.

His example: Queen Isabella funding Christopher Columbus. The pitch was essentially “I’m going to find a new route to India, there’s maybe a 50-60% chance you’ll never hear from me again, and by the way, maybe I just run away with the money.” The idea was wrong, Columbus discovered a completely different continent, didn’t even believe it himself, and Spain may not have been the main long-term beneficiary. But the bet paid off in ways no one could have predicted.

The whaling industry, the movie industry, book publishing, political campaigns: all follow this same pattern of backing long shots with uncertain outcomes.

Marc’s point about being “the last job”: if an AI could do all this, the talent-picking, the psychology, the navigation of a long and twisting path, he’d happily “hit the beach.” But the intangibles seem to be becoming more important, not less.

Why Zoomers Are the Best Founders Yet

Marc saves his most enthusiastic take for last: he’s “psyched for the Zoomers.” He describes 2015-2024 as a “very, very strange period” where things got weird, and Zoomers were on the receiving end.

His assessment of Zoomer founders: they don’t walk around feeling guilty about everything. They don’t deny wanting to be successful. They have no “moral hair shirt.” They’re original thinkers, incredibly well-trained because they grew up online watching thousands of hours of YouTube from great builders. They’re AI-native. They’re forceful, determined, and completely unapologetic.

Ben adds a more pointed observation: he’s never heard a Zoomer say “I’m going to do well by doing good.” Both he and Marc seem visibly relieved.

Marc compares it to Gen X as a reaction to Boomers: Zoomers are the reaction to millennials. If he had total control of his time, it would be “100% spent with Zoomers.”

Some Thoughts

The core thesis here is deceptively simple: market sizing is broken because it assumes static supply. But Marc and Ben have been living this thesis for 16 years now, and the examples keep compiling. The meta-insight is that this applies not just to individual companies but to venture capital itself: a16z’s own market (the number of fundable breakthrough ideas) is expanding along the same supply-driven curve.

A few things worth sitting with:

• The confidence framework for founders is more nuanced than the usual “we add value” VC pitch. Ben’s point about the vicious confidence cycle, where bad advice and inaccessible networks compound into slower decisions, rings true.
• The mythical man-month being dead is a genuinely significant claim. If you can now throw money at technological leads, the competitive dynamics of the entire software industry change.
• Marc’s enthusiasm about Zoomers is interesting not just as generational commentary but as an investing thesis: if the best founders are coming from a generation that’s AI-native, unapologetic, and online-educated, that changes how you source and evaluate deals.
• The Queen Isabella analogy is charming but also reveals something about how Marc and Ben see their role: they’re not optimizing a spreadsheet, they’re backing dreams with asymmetric payoffs, and the art of picking people may never be automatable.