
Podcast Notes | Conversations with Three Core Partners at a16z: Crypto Funds Must Be Restructured, Doubling Down on "Boundless Ambition" in the AGI Era
Organized & Compiled: TechFlow

TechFlow Summary: In this podcast episode, top Silicon Valley venture capital firm a16z officially announced the restructuring and expansion of its originally blockchain-focused a16z crypto fund into a brand-new frontier exploration fund, 「a16z Frontier」. Led by Chris Dixon, the fund will expand its investment scope from purely on-chain infrastructure to broader hardcore physical and digital frontiers including hardware, embodied robots, entertainment/gaming, space biology, and decentralized open-source intelligence. a16z co-founders Marc Andreessen, Ben Horowitz, and Chris Dixon appear together in an in-depth dialogue covering the evolution of technology cycles, power restructuring in the AI era, and epoch-making opportunities for entrepreneurs.
In the interview, these three godfather-level figures of Silicon Valley pointed out that the world is currently at a critical historical moment where the cognitive gap between early tech explorers and the general public is the widest ever. They explicitly refuted the pessimistic view that 「AI inevitably trends toward centralization」, profoundly analyzed the core value of open-source AI and cryptographic protocols in combating monopoly, stated outright that the commercial contract of traditional internet protocols has been broken by giants, and called on the new generation of entrepreneurs to break free from the trap of mediocre minor feature optimizations, bravely ask 「Is my vision ambitious enough?」, and become true builders in the untamed frontier where software and the physical world intersect.
Guests: Marc Andreessen (Co-founder & General Partner, a16z), Ben Horowitz (Co-founder & General Partner, a16z), Chris Dixon (Founding Partner, a16z Frontier / a16z crypto)
Host: Erik Torenberg (Well-known tech investor, Founder of Turpentine, Special Host for a16z Podcast)
Podcast Source: a16z Official Special Episode (https://www.youtube.com/watch?v=-9XQMHhpypc)
Original Title: The Frontier Ideas a16z Wants to Fund Now | Marc Andreessen, Ben Horowitz & Chris Dixon
Publication Date: October 22, 2026
Key Takeaways
- a16z crypto Fully Evolves into a16z Frontier: The original crypto fund has been officially expanded into a broader frontier tech fund, no longer limited to a single blockchain narrative, but comprehensively funding hardcore pioneers at the outermost edges of technological possibility, covering hardware, robotics, entertainment media, and bioscience.
- Historically Wide Cognitive Gap Between Early Explorers and the Masses: Ordinary people still view AI as simple conversational toys, while true early builders have fully stepped into the era of autonomous agents, local large model clusters, and super productivity. This cognitive potential difference harbors enormous dividends.
- Escaping the Mediocre Trap of the "Product Manager Brain": Marc Andreessen sharply criticized Silicon Valley's over-reliance on incremental improvement mindsets focused on trivial details like A/B testing and conversion rate tuning, emphasizing that only non-consensus leapfrog innovation and profound artistic taste can birth true platform-level breakthroughs.
- Open-Source AI Is the Ultimate Card Against Monopolistic Centralization: In response to Peter Thiel's past adage that 「crypto is libertarianism, AI is communism」, Andreessen and Dixon firmly believe that the combination of open-source large models and cryptographic ledgers is the only line of defense preventing tech giants and bureaucratic regulators from colluding to rent-seek.
- The Underlying Contract of the Internet Is Fracturing: The traditional model of 「creators contribute content, platforms reward with traffic and revenue sharing」 has been completely shattered by AI crawling and closed-loop summarization. Building new economic protocols featuring on-chain verification, provable ownership, and digital scarcity is urgent.
- New Forms of Physical Computing Are Brewing an iPhone-Level Explosion: Smart glasses, ambient spatial computing, and companion AI are reshaping human-computer interaction interfaces, coming to an end the absolute monopolistic status of smartphones spanning nearly two decades.
- Core Advice for Contemporary Youth: Reject step-by-step credentialism and mediocre career paths, deeply study physics, mathematics, and underlying systems engineering, master AI tools to multiply individual productivity a hundredfold, and become super builders endowed with both geek willpower and aesthetic taste.
II. Highlighted Quotes
- "The question the new generation of founders should ask themselves most is no longer 'Am I too ambitious?' but 'Am I ambitious enough?'"
- "Today, the gap between the technological possibilities understood by early adopters and those perceived by the general public and traditional management is the widest in the history of technology."
- "We have witnessed too many mediocre projects hijacked by the so-called 'Product Manager Brain', who excel at boosting conversion rates by one-thousandth but know nothing about leapfrog innovation that opens up entirely new dimensions."
- "Open-source is not mere charity; it is the most powerful anti-monopoly weapon in the history of human technology. When top-tier open-source models run freely across millions of edge devices worldwide, the blockade of centralizers is doomed to fail."
- "The initial economic contract of the internet has already been unilaterally torn apart by giants. You contribute all your knowledge, AI uses it for training, yet no longer returns even a single independent click. We must rebuild this system using cryptography and on-chain ownership."
- "Do not learn standardized skills that are about to be thoroughly automated by algorithms. Delve into hardcore physics, the essence of mathematics, and complex underlying systems, and become a person equipped with first-principles judgment."
III. Main Body
1. Expanding a16z Frontier: From a Single Crypto Track to Boundless Hard-Tech Frontiers
Host Erik Torenberg: Today, a16z made a major strategic announcement: officially expanding its globally renowned a16z crypto fund into a new frontier exploration fund, 「a16z Frontier」. Chris, what is the core thinking behind this?
Chris Dixon: Over the past few years, our team has been deeply contemplating the evolution of frontier technology cycles. When we first established the crypto-specific fund in 2018, it was the most exciting paradigm shift on the technological frontier at the time. But today, driven by the explosion of generative AI, foundational compute power, and modern cryptography, the frontier boundaries of technology have undergone massive spillover and convergence.
We see that the best founders no longer confine themselves to a single isolated label. They are building decentralized networks while simultaneously developing embodied robots, local agents, novel compute clusters, synthetic biology, and unprecedented immersive media. To comprehensively support these founders daring to chart the uncharted depths, we decided to radically broaden the fund's investment scope. Starting today, beyond just blockchain, every domain at the edge of human technological possibility will receive full backing through a16z Frontier. Founders should no longer ask if their ideas are too radical, but rather whether their visions are grand enough.
Marc Andreessen: Let me add some historical context. For fifteen years since a16z was founded, every time we faced a massive shift in technological waves, we proactively evolved our organizational structure and investment strategy. From initial cloud computing and mobile internet to later enterprise services, life sciences, and encrypted networks, technology has never evolved in isolation; it always presents network-like chain reactions. Now, AI is pulling forward the accumulated hardware, physics, and software advancements of the past decades. Frontier represents a mindset that is never satisfied with the status quo and continuously explores new continents.
Ben Horowitz: When we observe top-tier entrepreneurs, true revolutions often occur in the gray zones where different disciplines collide. If a startup is building humanoid robots powered by open-source large models while simultaneously distributing on-chain data rights and micro-payment rewards, you can hardly categorize it purely as AI or Crypto. The establishment of a16z Frontier aims to shatter all artificially drawn track barriers. As long as you are exploring truly unknown boundaries, here lies our most steadfast backing.
2. Is "Too Early" Still a Fatal Risk Today?
Host Erik Torenberg: There is a canonical saying in the VC world: 「In tech, being too early is almost indistinguishable from being completely wrong.」 Marc, you pioneered the Netscape browser in the 1990s and are widely recognized as one of the very few individuals who has repeatedly landed ahead of the curve multiple times. In today's ultra-rapid iteration cycle, is 「too early」 still a danger?
Marc Andreessen: Indeed, it is a classic warning because for decades, if your business idea led the era, you had to painfully wait for external supporting infrastructure to mature. For instance, twenty years ago if you wanted to build online streaming video, homes across America were still using dial-up internet. Without fiber broadband, smartphones, or cloud-native data centers, your company would exhaust its cash flow before the infrastructure caught up.
But something incredibly staggering has happened today: the timeline for foundational technological evolution has been compressed entirely. Current startups no longer need to lay cables from scratch or build semiconductor fabs themselves; global public clouds, robust open-source foundation models, and high-density mobile terminals are ubiquitous. More importantly, AI itself is a super accelerator that compresses code writing, prototyping, molecular simulation, and hardware rendering—tasks that once took years—down to days or even hours. Therefore, today, the temporal chasm caused by being 「too early」 is greatly bridged. Conversely, the real current danger is 「not being bold enough」. If you only launch a product slightly ahead of the curve by six months, tech giants and the open-source community will swamp you before you even officially go live.
Chris Dixon: I completely agree with Marc's assessment. Ten years ago, many said smart contracts and decentralized compute were too early. But precisely because of that foundational accumulation back then, we now have today's censorship-resistant, high-throughput distributed settlement networks. Within the Frontier umbrella, we are willing to take the risk of being 「too early」, because true excess returns and world-changing mega-corporations are always born in the stages when mainstream consensus is still bewildered and skeptical.
3. Historically Massive Cognitive Gap: Disconnect Between Early Explorers and Mainstream Public
Host Erik Torenberg: You've frequently mentioned in recent public interviews and articles that an extremely rare cognitive rift exists in today's tech landscape. How wide is the fault line you've observed specifically?
Chris Dixon: I believe this is the widest chasm between early adopters and public cognition since the dawn of the computer industry. Look around: mainstream society, executives of traditional multinational corporations, and even most regulatory officials still understand AI merely as an intelligent chat box that occasionally drafts official documents or tweaks resumes. They view it simply as another office efficiency plugin.
However, if you dive into the development sites in San Francisco or among the world's top geeks, you'll find that early builders have long entered a completely different universe. They collaboratively write complex operating systems through hundreds or thousands of autonomous AI agents, robots complete reinforcement learning in simulated environments at tens of thousands of times the speed of humans, and synthetic biology researchers use deep neural networks to design highly active proteins nonexistent in nature within weeks. This cognitive disparity is exponential.
Marc Andreessen: This mirrors exactly how ordinary people viewed the internet in the early 1990s. At the time, the vast majority considered webpages merely as electronic newspapers, utterly unable to imagine it would reshape global finance, supply chains, and geopolitics. This colossal cognitive potential difference is precisely the most disruptive moat for entrepreneurs. While incumbent giants still scrutinize new phenomena with old-system thinking, true builders have already reconstructed the entire foundation using new tools.
4. The Great Debate on Profit Distribution: Does the AI Dividend Belong to Incumbent Giants or Startups?
Host Erik Torenberg: The fiercest debate on Wall Street and in the VC community over the past two years concerns whether the massive value created by AI will ultimately be swallowed by existing tech giants or spawn a fresh batch of trillion-dollar startups. Ben, having experienced all cycles from mainframes, PC, internet, to mobile, how do you view this round of profit distribution?
Ben Horowitz: So far, on the surface, it is indeed giants like Microsoft, Google, Amazon, and Nvidia that hold the first-mover advantage, thanks to their ready-made massive distribution networks, compute budgets, and enormous enterprise customer bases. Plugging existing AI models into existing Office suites or search engines is the quickest path to monetization.
But tech history tells us a cruel iron rule: giants are forever constrained by the 「innovator's dilemma」. Their organizational structures, sales channels, and business models are tailor-made for the old world. Once a technological paradigm emerges that requires destroying their own core profit sources, giants instinctively hesitate and compromise. Take the early internet as an example: traditional telecom companies of that era owned all the fiber optics and telephone networks, and they ought to have dominated the internet, but eventually, Netscape, Amazon, Yahoo, and Google emerged victorious. Because new species never merely tweak old businesses by ten percent; instead, they are built from day one with entirely new business models. Startups carry no baggage, which is the fatal killer move.
Chris Dixon: This is especially true in the Frontier sector. If the future internet involves a cluster of autonomous agents conducting economic settlements and resource scheduling among themselves, the existing centralized business models based on credit card authorizations and advertising clicks will become completely obsolete. Agents do not need to view banner ads; what they need are frictionless micro-payment protocols and on-chain cryptographic credit. Only startups can engineer this qualitative leap in foundational business logic.
5. The Agent Economic Network: When Half the Residents of the Internet Are No Longer Human
Host Erik Torenberg: Following up on Chris's mention of agents, you have depicted a future of the 「Agent Internet」 in your investment thesis. In this network, how will AI entities interact and collaborate?
Chris Dixon: This is a fascinating and imminent evolution. Within the next three to five years, among the active interactive entities on the internet, humans might even retreat to a minority, with tens of millions of AI agents executing asset allocation, data retrieval, code deployment, and supply chain bidding on behalf of humanity.
But this raises a fundamental engineering and financial conundrum: How do agents establish their identities? How do agents complete cross-border settlements without traditional bank accounts? The traditional centralized banking system will never issue debit cards to lines of autonomous code. This is precisely why crypto protocols and decentralized agents are two sides of the same coin. Agents inherently need to hold their own cryptographic private keys, possess their own on-chain wallets, and achieve collaboration and settlement via smart contracts in permissionless networks.
Marc Andreessen: Incredible changes will even occur at the social network level. Social platforms of the past two decades were peer-to-peer follows between humans, whereas future social networks may become highly mixed ecosystems. Agents will possess their own personalities, knowledge reserves, and economic demands, engaging in viewpoint debates, content creation, and even forming their own decentralized collaborative organizations in social squares. This will completely reconstruct our perception of 「communities」 and 「media」.
6. Breaking Free from the "Product Manager Brain": Artistic Breakthroughs in Gaming, Film, and Physical Hardware
Host Erik Torenberg: Marc, you recently put forward a razor-sharp viewpoint, sharply criticizing the prevalence of the 「Product Manager Brain」 in Silicon Valley in recent years. What negative impacts do you believe this mindset has brought to contemporary tech innovation?
Marc Andreessen: The so-called 「Product Manager Brain」 is a highly harmful bureaucratic mindset evolved from the mature phase of the mobile internet. These individuals have undergone rigorous metric training in large tech corporations, believing that the essence of all innovation is monitoring retention curves, running multivariate A/B tests, and shifting button colors from dark blue to light blue in exchange for a one-per-thousand click-rate boost. They reduce business to cold, hard conversion formulas.
In mature software markets where growth has peaked, this kind of refined operations might squeeze out a bit of profit. But in the face of true frontier exploration, the 「Product Manager Brain」 is not only useless but could even strangle the sprouts of all great innovations. Imagine if Steve Jobs conducted focus groups and A/B tests while developing the iPhone; he would only have ended up with a BlackBerry featuring a smaller keyboard.
Marc Andreessen: True breakthroughs rely on deep taste, artistic intuition, and unyielding engineering obsession at all costs. Especially in AAA gaming, cinematic storytelling, and entirely new hardware terminals, masterpieces are always sculpted by creators wielding strong personal will and extraordinary taste. The true significance of AI is not to help product managers generate a few more dashboards, but to place the most top-tier generative tools into the hands of small teams possessing extreme taste. In the future, an independent studio of just ten people will have the capability to produce interactive universes that surpass Hollywood's top blockbusters in visual fidelity and worldview complexity.
Ben Horowitz: I couldn't agree more. The essence of business has never been algorithmic self-recursion, but rather touching upon the deepest emotions and curiosity of humanity. Entrepreneurs must reclaim the founder's will, dare to make heavy bets on unknown experiences, instead of handing decision-making power over to mechanical dashboards.
7. Will AI Trend Toward Centralization? Why Open-Source Models Concern Civilizational Future
Host Erik Torenberg: Well-known Silicon Valley investor Peter Thiel coined a widely circulated aphorism about a decade ago: 「Crypto is libertarianism, AI is communism.」 His core logic is that blockchain decentralizes power, whereas AI naturally tends toward the massive computational power and data monopolies of centralization. Marc, Chris, how do you view this assertion?
Marc Andreessen: Peter's observation at that time had its acuteness in a specific historical phase, because early deep learning indeed heavily relied on centralized data centers and servers of oligarchic giants. If you only have closed-source models, and only two or three multinational giants and governments worldwide can control the smartest machine brains, then AI would indeed evolve into the most terrifying tool of centralization in human history.
But this is precisely why we must defend open-source AI to the death! The rise of open-source large models has completely shattered this deterministic fate of centralization. From Meta open-sourcing the Llama series to globally active open-source communities, deeply optimized architectures, and quantization algorithms, we witness state-of-the-art intelligence being rapidly compressed. Today, a model rivaling supercomputers from a few years ago can run smoothly directly on your laptop, or even local chips in smartphones. Open-source means code is public to all of humanity, weights belong to no bureaucratic review agency, and anyone can freely fork, fine-tune, and deploy.
Chris Dixon: This aligns perfectly with the spirit of cryptography. Cryptography destroyed the privilege of financial intermediaries, while open-source models destroy the intermediaries of thought and intelligence. In Washington and other regulatory hubs, certain vested-interest giants are cleverly lobbying under the guise of 「security」, attempting to define open-source software as illegal, thereby building compliance barriers to maintain their monopolistic status. This is a classic 「Baptist and bootlegger」 political alliance. Our support for a16z Frontier is to unswervingly invest in open-source ecosystems, decentralized compute, and open protocols, absolutely refusing to let the future of technology be privatized by a handful of giants.
8. The Collapse of the Internet's Underlying Contract: Trust Crisis Between Creators and Platforms
Host Erik Torenberg: Chris, you profoundly discussed the collapse of internet protocol contracts in your book 《Read Write Own》. In the AI era, is this contractual breakdown further intensifying?
Chris Dixon: The break has become completely overt. Early Web2 platforms were built on a fragile tacit understanding: independent creators, blog authors, news outlets, and community volunteers contributed massive amounts of high-quality content for free, and in return, platforms offered them opportunities for distribution traffic, audience building, and commercial revenue.
But once giants achieved monopoly over network effects, the first thing they did was close the gates and beat the dogs, constantly tightening algorithms and forcing creators to pay for traffic. And now, entering the generative AI era, large model companies scrape the entirety of humanity's decades-long efforts across the entire web, then directly serve final answers on search engines and Q&A interfaces, no longer diverting even a single independent visit to the original websites. For independent creators, the original economic cycle has been completely severed.
If this continues unabated, humanity's high-quality original content will dry up, and the internet will be flooded with low-quality garbage generated by machines talking to themselves. The only way to break through is to construct verifiable data ownership through blockchain protocols. Through on-chain rights confirmation, enable every call, inference, and training of data to be traced and automatically earn economic returns. Reshaping this contract is one of the most crucial missions of the Frontier fund.
9. Farewell to Smartphone Hegemony: What Will Take Over the Post-iPhone Hardware Era?
Host Erik Torenberg: Since the original iPhone launched in 2007, smartphones have dominated human digital life for nearly two decades. Looking toward the next decade, what form of physical computing carrier can truly break the hegemony of this glass screen?
Ben Horowitz: Smartphones are indeed a perfect industrial miracle, but their interaction format is essentially counter-intuitive. Humans must bow their heads, detach attention from the real world, and stare fixedly at that five-to-six-inch piece of glass in their palms. This is not the ultimate form of coexistence between technology and humanity.
What we currently favor most is all-day wearable computing devices, particularly smart glasses and environment-aware hardware. When you wear a device weighing no different from regular glasses, equipped with a front-facing high-dynamic-range camera and directional audio microphone, it means you possess a companion AI capable of sharing your first-person perspective and completely understanding the physical environment. It knows what you are looking at, understands who you are speaking to, and provides auxiliary information right in your ear at the perfect moment, without requiring you to pull out your phone and slide to unlock.
Marc Andreessen: This is a fundamental leap from 「tool-based computing」 to 「environmental awareness intelligence」. Spatial audio, retinal projection, and haptic feedback are maturing. Future hardware will be not merely an information receiver, but an external extension of human cognitive capabilities. This track is birthing a brand-new hardware supply chain ecosystem worth hundreds of billions of dollars.
10. If You Were Exactly Eighteen Today: Survival Rules for Youth from Silicon Valley Godfathers
Host Erik Torenberg: At the end of this interview, I'd like to pose a question on behalf of all curious and confused younger generations: If you were exactly eighteen today, preparing to enter university or step into society for the first time, what fields would you choose to delve into?
Marc Andreessen: My first piece of advice is: stay away from mediocre, smooth trajectories. Absolutely do not study standardized majors designed to train you into a cog. If you go learn administrative skills that handle documents according to fixed regulations, you will find yourself completely replaced by large models before you even graduate.
Return to the physical world and first principles. Deeply study foundational mathematics, theoretical physics, underlying computer systems engineering, and chemical materials. Laws of physics are the hardest truths in the universe; they never lie. Meanwhile, be extremely hungry to master the cutting-edge AI tools. Eighteen-year-olds today possess superpowers unprecedented in history. A young person proficient in modern generative tools and programming agents is worth as much as an entire software development department of twenty engineers used to have. Go build tangible things, launch real products; code and user feedback trump all prestigious diplomas.
Ben Horowitz: I would add something regarding character and willpower. No matter how technology changes, the scarcest resource in the business world will always be genuine leadership and accountability. Explore those dangerous fringes amidst stormy winds, learn to face failure, learn to organize eccentric geniuses with shared ideals, learn to hold firm to your intuition when the masses sneer at you. True wealth in the future will forever belong to those builders who roll up their sleeves and sweat in workshops and code repositories.
Chris Dixon: Maintain a geek spirit, remain skeptical of all overly packaged authoritative consensus. Go join the newly opened continents currently under pioneering, even if they look clumsy and rough today. Twenty years ago, it was open-source software; ten years ago, it was blockchain; today, it is the hard-tech and agent world exemplified by a16z Frontier. Brave steps onto the frontier, for the future has only just begun.















