Jun 19, 2024 · 45m · a16z
Marc Andreessen & Andrew Chen Talk Creative Computers
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this a16z Podcast episode, venture capitalist Marc Andreessen and host Andrew Chen discuss techno-optimism, regulatory challenges, artificial intelligence, and the evolutionary power of the gaming industry. They examine how probabilistic computing and game design methodologies are driving the next fundamental platform shifts for technology founders.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The host as informed peer 3.2 Guest teaching 5.3 Guest disagreement 2.3 The host pushing back 1.3
speaking balance: gold is the host, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 18:15 Slamming Big Tech Regulatory Capture
Marc aggressively attacks incumbent tech giants, calling out their aggressive Washington lobbying for complex rules designed to entrench themselves while locking out startups.
Hardest push from the host ▶ 8:57 Pushing Back on Blaming MillennialsAndrew immediately interrupts Marc's joke about millennials causing tech self-hatred, asserting that Gen Z will save us and steering the conversation back.
Biggest teaching moment ▶ 25:59 Deterministic vs Probabilistic Computing MasterclassMarc reframes the entire AI wave away from standard software shifts into a foundational historical shift from 80 years of literal Von Neumann computing to probabilistic neural networks.
The host holds their own ▶ 19:00 Mapping Regulatory and Geopolitical Realities to GamingAndrew demonstrates deep industry expertise by connecting UGC platforms, European safety laws, and geopolitical tensions around Tencent directly into Marc's regulatory framework.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The host as informed peer | Guest teaching | Guest disagreement | The host pushing back | Why |
|---|---|---|---|---|---|---|
| Personal Connection to Gaming and Parenting Insights | 3 | 4 | 1 | 1 | Andrew Chen introduces gaming topics and offers industry insights around VR retention data. Marc Andreessen shares parenting experiences and rich historical context on early gaming breakthroughs like Space War and Pong. | |
| The Techno-Optimist Manifesto and Historical Moral Panics | 2 | 6 | 3 | 1 | Andrew Chen asks Marc to explain the root causes of technology moral panics. Marc delivers a wide-ranging historical overview covering 19th-century bicycle face fears, 1990s video game panics, and modern AI/crypto doom-mongering. | |
| Generational Dynamics and Digital Media Literacy | 3 | 5 | 2 | 2 | Andrew queries why tech insiders themselves adopt doom narratives. Marc jokingly points at millennials before breaking down generational psychology and mimetic defenses, while Andrew counters that Gen Z will save us and highlights subcultural diversity. | |
| Government Bureaucracy, Regulation, and Historical Innovation | 3 | 6 | 3 | 1 | Andrew asks about pro-tech policy and government innovation projects. Marc explains adverse selection in government, contrasting top-down projects like Apollo with modern high-speed rail failures and European versus American regulatory models. | |
| Political Engagement for Founders and Fighting Regulatory Capture | 2 | 5 | 4 | 1 | Andrew asks how startup founders can actionably support tech-optimism beyond social media posts. Marc strongly criticizes Big Tech incumbents for lobbying for regulatory capture and advises startups to build policy teams and political political action funds. | |
| Game Industry Realities and Public Narrative Volatility | 4 | 5 | 3 | 2 | Andrew outlines how user-generated gaming platforms face immediate regulatory and geopolitical scrutiny. Marc agrees and illustrates media narrative volatility by citing early dismissals of Facebook's $1B offer before it was declared an evil threat. | |
| Speculating on a New AI Medium and Creative Computers | 4 | 4 | 1 | 1 | Andrew prompts Marc on what new venture he would build if exiled. Marc envisions a novel AI-generated dream medium, which Andrew builds upon using historical parallels from YouTube's native video formats. | |
| Deterministic versus Probabilistic Computing Models | 2 | 7 | 2 | 1 | Andrew asks how AI compares to prior computing waves. Marc details a foundational shift from 80 years of Von Neumann deterministic computing to probabilistic, creative computing architectures. | |
| Analyzing Platform Shifts: Incumbents versus Startups | 4 | 6 | 2 | 2 | Andrew explores whether AI incumbents hold permanent structural advantages over startups. Marc breaks down VC platform shift dynamics, comparing incremental feature additions to total paradigm shifts like Midjourney replacing manual pixel editing. | |
| The Case for Open-Source AI and Edge Deployment | 3 | 6 | 4 | 1 | Andrew introduces open-source AI models versus centralized cloud architectures. Marc vigorously defends open source and edge computing against incumbent calls to regulate or ban open weights. | |
| The Evolution of Gaming Business Models and Founders | 4 | 5 | 1 | 1 | Andrew asks about a16z's thesis for dedicated gaming investments. Marc explains how gaming evolved from hit-driven media into 30-year platforms, supported by Andrew sharing metrics from the firm's Speedrun accelerator. | |
| Applying Gaming Skill Sets to Education and Healthcare | 4 | 5 | 1 | 1 | Marc discusses how game mechanics can solve metabolic health and education problems. Andrew reinforces the idea by contrasting the $1B spent on Grand Theft Auto with modest investments in computer science education. |
Statements from this episode (30)
Andreessen: Major platform shifts demand reinventing products from scratch
“Every once in a while you get a platform shift that's like dramatic enough where it basically means that you both kind of can and have to reinvent from scratch the whole idea of what the thing is in the first place.”
Andreessen: Kerbal Space Program is exceptionally effective for educational subjects like physics
“And by the way, we were homeschooling, and so he's doing, like, a ton of, like, along with this, like, a lot of math and physics and computer science and rocket design, you know, with his science teacher and stuff like that. So, like, Kerbal Space Program turn…”
Andreessen: Gaming Will Continue Driving the Scale of Emerging Tech like AI
“And I fully expect this trend to continue, and it's obviously what we're seeing with AI right now.”
Andreessen: Games Are the Only Known Killer Apps for VR and AR
“Both VR and AR, like, it's crystal clear. The known killer apps right now are games, like, by a very wide margin.”
Andreessen: US violent crime dropped as video games became popular
“Video games took off and actually, you know, for the next 30 years actual, you know, violent crime in the US actually dropped like a rock.”
Andreessen: Pessimism sounds deceptively sophisticated compared to optimism
“It always sounds more sophisticated to be negative than positive, right? It's like the positive person gets accused of being Pollyannish and wide eyed and, you know, immature and, you know, not cynical enough. And then the pessimistic person always sounds worl…”
Andreessen: Future technologies will prove beneficial on net despite panics
“There's just such a rich history of the moral panic predicted societal doom, and then it actually turned out everything, you know, overwhelmingly on net was fine, and I think that will continue to be what happens.”
Andreessen: Boomers are most vulnerable to social media manipulation
“The person most prone to be really screwed around with social media would seem to me to be somebody with no mimetic defenses at all, which is like a boomer, right? And so, you take, like, somebody raised on, like, Walter Cronkite, and you, like, inject them on…”
Andreessen: Younger generations will navigate digital media better than elders
“New generations are going to be like much better at dealing with this whole media politics, you know, kind of matrix.”
Andreessen: Top talent capable of mega-projects moved to private sector
“The kinds of people who can run projects like that just don't work for the government anymore. They, they're in the private sector.”
Andreessen: California high-speed rail spent $10B with zero ground broken
“The California high-speed rail project. Which is now ten billion, it's over ten billion in spend with zero ground broken.”
Andreessen: Nuclear Regulatory Commission went 40 years without approving new designs
“They did not approve a new nuclear power plant designed for 40 years.”
Andreessen: a16z launched an ambitious political program to defend startups
“We've launched a sort of a really ambitious political program in the last year, because we've just had it with people basically taking free shots on goal on startups, and so we're trying to support a lot of the politicians on the right side of these issues”
Andreessen: Young tech startups will need to build political programs and PACs
“I mean, I think more, more companies, especially more young companies in our industry are going to have to step up and like build policy groups, policy teams and political programs and have packs and, you know, have put real effort into this.”
Andreessen: Big Tech is aggressively lobbying for regulatory capture against startups
“Quite honestly, I mean, part of what's happening is big tech is against us. There, there's a, there's basically a lot of the political issues now are a civil war inside the tech industry where, The big tech companies are lobbying very aggressively for what the…”
Chen: User-generated gaming platforms face same regulatory scrutiny as social media
“A lot of the most interesting new platforms like you know, UEFN and Roblox, which are user generated, you know, content platforms, you know, all the metaverse work that's also happening as well. You know, all of those kind of fall into user generate, you know,…”
Andreessen: Journalist advised Zuckerberg to accept $1B Yahoo acquisition offer
“Facebook got just tons of, you know, it's just like this is the most trivial, stupid, useless, you know, thing that doesn't matter, you know, Facebook, there was almost a deal for Facebook to get bought by Yahoo early on, and I remember a certain very prominen…”
Andreessen: Media narratives about tech jump directly from useless to evil
“And then basically within like five years, it was like Facebook is like the living demon incarnate on earth and destroying everything. Right? It went from like useless to evil in one step. Right? You know, the headline never gets never just like, oh, this is k…”
Andreessen: Legacy movie and game studios won't create new AI media
“The movie studios aren't going to do it because they just make movies. The game studio companies aren't going to do it because they make games.”
Andreessen: AI Represents a New Paradigm of Probabilistic Computing
“When I look at AI, what I see is sort of a new kind of computer, like a fundamentally new kind of computer... It's a probabilistic computer and you ask the same question twice, it gives you different answers, which is just in and of itself a pretty mind blowin…”
Andreessen opposes locking down AI models, cheering on open jailbreaks
“The last thing I want is these things locked down. Like, I want these things to be out there, like, doing all kinds of crazy stuff.”
Andreessen: Every microchip will run AI alongside cloud god models
“Basically, I think anything with a chip is gonna have an AI, and you're just gonna, you just basically, you're just gonna assume that everything has an AI and there's gonna be, you know, many billions of tiny models running everywhere. There's gonna be a few b…”
Andreessen: Generative AI prompts will render manual pixel editing obsolete
“Because you'll live in this new world, and in the new world, you'll just tell the machine what you want for a photo, for an image, and it'll generate you as many options as you want, and it's generating you something that's better than you would have been able…”
Andreessen: a16z's mistakes stem from overestimating platform turnover speed
“When we make mistakes, it's Because we kind of overestimate the degree to which this is going to happen. But our whole thing is to basically try to get in front of this and try to, you know, basically make the assumption that you're going to have complete turn…”
Andreessen: Local hardware devices like doorknobs will run onboard LLMs
“There's just gonna be such a giant number of like local use cases. Like, you know, your doorknob is gonna have an LLM in it that's gonna control whether or not people get through the door, right?”
Andreessen: Static video will become obsolete like silent movies within a decade
“You could imagine looking back in a decade and being like, I can't believe people used to watch, you know, static videos at all. Like, it just doesn't make any sense. It's like watching a silent movie. Obviously, videos should be, like, responsive in real time”
Andreessen: Modern games serve as permanent platforms lasting decades
“Game as sort of permanent platform and, ah, you know, sort of foundation that then you end up building 20 or 30 or 40 years of experiences on top of.”
Andreessen: Future Great Educational Experiences Will Feel Like Games
“Someday there's going to be a great educational experience and it's going to look and feel like a game. Like, that just seems, like, crystal clear.”