a16z removed Crypto from its fund name, but that doesn't mean it's preparing to leave the crypto market.
On October 22, Andreessen Horowitz (a16z) announced that it would rename its crypto fund to a16z Frontier. The fund is still led by Chris Dixon, with a16z founders Marc Andreessen and Ben Horowitz, as well as partners Eddy Lazzarin and Guy Wuollet, jointly participating in management.
The official explanation is clear: this adjustment is about expanding the investment scope, not exiting crypto. The new fund will continue to invest in crypto and blockchain, while also covering more new technologies and new markets that take years to mature.
What is truly worth paying attention to is not the missing word in the name, but that a16z is redefining its investment targets. From internet software to blockchain, and then to AI, hardware-software integrated products, and the economic networks between machines, what it wants to find has always been the technologies that may give rise to the next generation of markets.
From internet entrepreneurs to a top Silicon Valley venture capital firm
In 2009, Marc Andreessen and Ben Horowitz founded Andreessen Horowitz, which is today's a16z.
Neither of them was simply an investor who transitioned from the financial industry. Andreessen had co-founded Netscape, and Horowitz had run the enterprise software company Opsware with him, which was eventually acquired by HP. They personally experienced the process of internet technology moving from early innovation to commercialization.
After that, a16z invested in a batch of technology companies such as Facebook, Airbnb, GitHub, Instagram, Slack, and Figma, covering social networks, consumer internet, developer tools, and enterprise software.
These companies had different businesses, but they shared one thing in common: they were not just optimizing old products, but building product forms and business models on new computing platforms.
This investment approach later extended to the crypto field. What a16z focuses on is not just asset prices, but whether new technologies can form platforms, developer tools, financial services, and applications for ordinary users.
It viewed crypto as the next-generation computing platform early on
a16z did not enter the market only after crypto became mainstream.
In 2018, a16z officially launched the independent a16z crypto fund, with an initial size of $350 million. The fund emphasized long-term investment, continued deployment across market cycles, and providing entrepreneurs with recruiting, regulatory, marketing, and operational support.
It invested not only in tokens, but also in projects spanning trading platforms, public blockchains, DeFi, NFTs, wallets, and cross-chain infrastructure, including Coinbase, Solana, Uniswap, OpenSea, Phantom, and LayerZero.
The key to this investment logic is that if blockchain can become new computing and financial infrastructure, then the opportunities will not exist only in token trading. Base-layer networks, developer tools, financial products, and user applications could all form independent commercial markets.
The establishment of a16z crypto was itself a bet. It sought to allocate in advance to networks and companies that might grow in the future, while the industry was still young and business models had not yet fully taken shape.
From $350 million to $4.5 billion, then back to $2.2 billion
The fundraising history of a16z crypto also records the capital cycle of the crypto industry.
· In 2018, the first fund raised $350 million.
· In 2020, the second fund raised $515 million.
· In 2021, the third fund raised $2.2 billion.
· In 2022, the fourth fund raised $4.5 billion.
· In 2026, the fifth fund raised $2.2 billion.
The $4.5 billion fund in 2022 was one of the largest crypto-focused venture capital funds at the time. Four years later, the size of the fifth fund fell back to $2.2 billion, less than half of the previous fund.
According to a May 2026 report by The Information, a16z crypto's five funds had cumulatively committed about $9.8 billion in capital.
However, a decline in fundraising size alone is not enough to prove that a16z is bearish on crypto. Fund size reflects both the manager's investment judgment and fundraising cycles, the pace of capital deployment, and market opportunities. The smaller new fund in 2026 at least shows that a16z did not simply continue the continuously expanding fundraising model of the previous round.
More notably, when the fifth fund was announced in May, a16z had already begun shifting the focus of discussion from price cycles to actual usage.
Months before the rebrand, it had already begun emphasizing crypto's "usage cycle"
In May 2026, when a16z crypto announced the raising of its fifth fund, it offered a judgment: speculative cycles bring attention and capital, some of which will be wasted, while another portion will be used to build the infrastructure that ultimately remains.
What the market really needs to observe is not just how much prices have risen, but how many people continue to use these products after the market cools.
At the time, a16z specifically emphasized several directions.
Stablecoins. Their use is extending from a medium of exchange within trading platforms to savings, cross-border remittances, and everyday payments.
Onchain finance. Perpetual contracts, prediction markets, and onchain lending are attempting to move more financial activity onchain.
Asset tokenization. Traditional assets have the opportunity to gain new issuance, settlement, and circulation methods through blockchain.
AI Agents. Software agents may in the future autonomously call upon computing power, obtain data, purchase services, and complete payments and transactions on behalf of users.
These directions appear scattered, yet behind them lies a common question: Can blockchain transform from a relatively independent crypto industry into broader financial and software infrastructure?
This is also the key to understanding a16z Frontier. Crypto remains an investment target, but it no longer needs to be an identity label for all related companies.
Why remove Crypto from the name?
The core reason Chris Dixon gave in the official announcement is that the scope of blockchain applications is expanding, and traditional industry classifications are increasingly unable to accurately describe a company's business.
When fintech companies use blockchain to provide payment services, traditional financial institutions launch tokenized assets, and software products begin integrating onchain settlement into their own operations, should all these companies be called "crypto companies"?
Dixon raised an example in the announcement: Is Stripe really a crypto company?
This question is not meant to label Stripe, but to illustrate that what a company does and what technology it uses are becoming two different things.
In the future, a company may use blockchain while positioning itself as a payments or fintech company; it may also combine AI, financial services, and crypto technology to create new products that had no corresponding category in the past.
If a fund's investment scope remains limited by the word Crypto, it may be difficult to accurately cover these crossover areas.
Therefore, a16z Frontier's adjustment is not just about expanding the industry list, but also an attempt to change the way investment targets are defined. It no longer revolves solely around a technology category, but seeks new companies that are spawned by new technologies and have not yet formed mature markets.
Frontier's keywords: exoplanets and second-order effects
In related interviews, a16z uses "inner planets" and "outer planets" to explain this investment approach.
Inner planets represent core areas that have already formed and are relatively easy to identify. Outer planets refer to ideas that are earlier, stranger, and harder to categorize.
Traditional investment teams can focus on areas with clear existing markets, but there will always be some technologies and entrepreneurs that fall outside existing categories. What Frontier wants to find is precisely these opportunities that have not yet formed mature tracks.
Another important concept is "second-order effects."
After the widespread adoption of smartphones, directly user-facing applications such as Instagram grew rapidly—this is the first-order effect brought by technological change. But the development of smartphones also drove advances in hardware capabilities such as sensors and chips, further creating conditions for drones, VR, and other products. These indirectly generated new markets are second-order effects.
If investors focus only on the popular applications first spawned by a technology, they may miss the next round of opportunities that emerge only after the technology matures.
In the interview, Ben Horowitz summarized this approach in one sentence: what they look for is not "good ideas that look like good ideas," but "good ideas that look like bad ideas."
This does not mean that the more absurd a project is, the more worth investing in it. The real difference lies in whether the entrepreneur has grasped some key condition that is not yet widely understood by the market and that can ultimately make a seemingly impractical idea work.
AI is widening the gap between technology pioneers and traditional enterprises
a16z's discussion of Frontier is also closely tied to the technological changes brought about by AI.
In an interview, Marc Andreessen stated that he believes the cognitive gap between current technology pioneers and traditional enterprises is the largest he has seen in his entire career, and it is still widening.
He and Horowitz were involved in founding Netscape in 1994. Andreessen even imagined that if today's AI tools had existed then, the two of them might have quickly completed work that previously required a massive team.
Such remarks carry obvious personal judgment, but they point to a practical investment question: when software development, mathematics, and engineering capabilities are amplified by AI tools, projects that were previously difficult to launch due to high costs or technical complexity may begin to become feasible.
In the past, investors might have worried that a project's ambition was too great. Horowitz's counter-question is whether, after AI changes development efficiency, a project's ambition is great enough?
If all teams can use AI to develop software faster, merely making existing products faster and cheaper may not necessarily create long-term differentiation. The greater opportunity may come from which things that were originally impossible become feasible for the first time because of AI.
When AI lowers R&D costs, can small teams challenge big companies' projects?
In the interview, a16z put forward a bold judgment: integrated software-hardware projects that previously might have required 1,000 people to complete could in the future be completed by teams of 20 to 30 people.
This is not to say that all hardware projects can immediately be reduced to a few dozen people. Constraints such as manufacturing, supply chains, funding, testing, and security validation will not automatically disappear because of AI.
The more realistic change is that AI may lower the costs of R&D, software engineering, hardware reverse engineering, and product iteration.
In the past, a software engineer who wanted to enter the hardware field might have been blocked by specialized tools, firmware, and complex engineering processes. AI-assisted code generation and reverse engineering have the potential to enable more software entrepreneurs to try drones, robots, smart devices, and products combining software and hardware.
If these tools continue to mature, investment opportunities will not only lie in how many more AI application companies emerge, but also in which industries will reopen because the R&D threshold has been lowered.
This is also the difference between Frontier and traditional industry classifications: what it wants to focus on is not an additional product in an existing market, but whether technological change makes a market that was originally difficult to establish become feasible.
AI Agents and Stablecoins May Meet in the Machine Economy
AI agents are another important direction a16z sees for connecting AI and crypto.
Future software agents will not only generate text but may also call software, fetch data, purchase computing power, use external services, and coordinate tasks across different systems.
The question that follows is: when software agents begin to act autonomously, how do they pay? How do they settle between different services? How do they complete high-frequency, small-value, or cross-border transactions without requiring humans to operate each one?
Stablecoins and programmable blockchains may become tools for solving these problems.
Compared with traditional payment systems, on-chain assets can be transferred across global networks and can also interact with other applications through programs. In theory, AI agents can use stablecoins to pay for computing power and data services, and can also settle between different agents.
In this vision, the role of crypto is no longer limited to human investors trading digital assets; it provides a set of infrastructure for economic activity between software agents.
However, this is still an emerging direction. Autonomous transactions, payment authorization, risk control, and liability attribution for AI agents all still have real-world problems to solve. Being technically capable of completing a transaction does not mean a sustainable machine economy has already formed.
For a16z, the appeal of such intersectional areas is that they are hard to simply classify as either AI or crypto, yet may depend on the development of both.
Open-source AI Is Also Part of This Investment Logic
In related interviews, a16z also discussed open-source models, technological transparency, and the concentration of power.
They worry that if AI capabilities and model control remain concentrated in the hands of a few closed-source labs or large institutions for a long time, ordinary developers, researchers, and startups will find it increasingly difficult to understand, improve, and audit this technology.
The value of open-source models is not only in lowering usage costs, but also in enabling more people to inspect models, conduct research, develop new applications, and participate in solving safety issues.
a16z specifically mentioned that open-source AI is developing rapidly in both the United States and China, and expressed its policy stance against excessive restrictions on open-source technology.
But open source does not equal inherently safe. Open models can also bring risks of misuse, and technological transparency cannot replace safety assessments. a16z's support for open source reflects its judgment on the path of technological development and the distribution of innovation power, rather than proof that open-source models have solved all safety problems.
From an investment perspective, the diffusion of open-source models could create multi-layered opportunities, including the models themselves, development tools, edge devices, software-hardware integrated products, and new business models formed around these technologies.
a16z Doesn't Want to Be Defined by a Single Track
From internet software to crypto networks, and then to AI and frontier hardware, a16z's investment history has always revolved around one question: What is the next important computing platform, and what companies that don't exist today will it spawn?
In the past, the name a16z crypto made it easy for the market to understand its investment boundaries. But as blockchain enters fintech, AI Agents begin to involve payments and settlement, and AI tools may change hardware R&D, the boundaries between technologies are becoming increasingly difficult to delineate.
Frontier attempts to shift the investment focus from "a certain technology" to "new markets created by new technologies."
This does not mean crypto has lost its importance. On the contrary, blockchain may still play a role in programmable money, machine-to-machine transactions, and the economic incentive mechanisms of new networks. What has changed is that it does not have to become the core identity of all related companies.
a16z has also made it clear that crypto investments and related regulatory initiatives will continue. Chris Dixon continues to lead the fund, indicating that this adjustment is about expanding scope on the basis of existing investments, rather than announcing the end of crypto operations.
From betting on internet startups in 2009, to launching an independent crypto fund in 2018, and then expanding the fund into Frontier in 2026, a16z has been constantly searching for new technology platforms and the markets they may spawn.
The significance of this renaming is not which word it deleted, but that it attempts to push investment boundaries beyond existing industry classifications.
As AI lowers R&D costs, blockchain provides programmable economic infrastructure, and software-hardware integrated products begin to open up to smaller teams, the next generation of important companies may be born in the intersections of existing tracks.
What a16z Frontier wants to do is find the people building these markets before they have been clearly named.
Welcome to join the official BlockBeats community:
Telegram Subscription Group: https://t.me/theblockbeats
Telegram Discussion Group: https://t.me/BlockBeats_App
Official Twitter Account: https://twitter.com/BlockBeatsAsia
