The internet has entered a new phase, an AI-driven internet era.

Organized & Compiled by: Deep Tide TechFlow

Introduction

a16z crypto recently released a year-end special program, divided into two parts. The first episode invited Sam Broner, Maggie Hsu, Daren Matsuoka, Joachim Neu, and Chris Lyons to discuss stablecoins, the crypto App Store, the current state of industry projects, the development of infrastructure, and prospects for 2025. The second episode featured Carra Wu, Eddy Lazzarin, and Karma as guests, delving into the hot topic of AI Agents, including the integration of AI and crypto, how we can effectively distinguish between humans and robots in the context of AI's growing prevalence, and discussions on decentralized, truly autonomous chatbots, emphasizing the potential autonomy and commercial freedom AI may achieve in the future.

Deep Tide TechFlow has transcribed and integrated these two episodes of a16z Crypto, and below is the complete dialogue.

Guests:

Sam Broner, Partner at a16z Crypto investment team;

Maggie Hsu, Partner at Andreessen Horowitz;

Daren Matsuoka, Partner at a16z Crypto investment team;

Joachim Neu, Researcher at a16z Crypto;

Chris Lyons, President of Web3 Media at a16z Crypto;

Carra Wu, Partner at a16z Crypto investment team;

Eddy Lazzarin, CTO at a16z Crypto;

Karma (Daniel Reynaud), Research Engineering Partner at a16z Crypto

Hosts: Robert Hackett & Sonal Chokshi

Podcast Source: a16zcrypto

Original Titles:

Talking trends 2025 (part 1): Stablecoins, app stores, UX, and more;

Talking trends 2025 (part 2): AI x crypto

Release Date: December 20, 2024

Part One

Stablecoins

Sonal: Sam, your "big idea" is about stablecoins; you've written a lot about it recently… and Robert and Daren have produced the "Crypto State Report," with their main conclusion being that stablecoins have found product-market fit (i.e., the market demand for stablecoins is highly aligned with product functionality). But what we really want to understand is, why now?

Sam:

Over the past year, the technological platform for stablecoins has significantly improved, with transaction costs dropping from $5 per transaction to less than a cent. This has greatly reduced payment costs—but retailers, merchants, and other businesses that could benefit the most have not yet adopted this technology on a large scale.

Many believe that the early adopters will be tech-centric businesses… but these businesses often have high profit margins and do not urgently need improvements in their cost structures. Therefore, those with lower profit margins—like corner stores, restaurants, and family-owned businesses—may be the ones most eager to accept stablecoin payments.

We're talking about businesses like coffee shops—currently with a profit margin of only 2%—that could double their profits through stablecoin payments. In fact, this would make businesses that are nearly unprofitable today moderately profitable, which is a huge change.

Sonal: One thing that really opened my eyes is when you mentioned that small businesses get nothing from credit card companies; they not only pay high fees but receive almost no return.

Sam:

Exactly! A significant feature of credit cards is that they provide consumers with fraud protection. This has been very helpful in driving online sales… but when you pay at a coffee shop, this protection is almost meaningless.

Each transaction incurs a fixed fee of $0.30, plus an additional 2% fee. This means that for a $1.50 cup of coffee, nearly $0.30—one-fifth of the money—goes into the pockets of the payment service provider, while they provide almost no actual value in this transaction.

This 2% fee is pure profit for the payment service provider, while it is a pure loss for the local coffee shop. I am very much looking forward to these small businesses reclaiming that $0.30/$0.35 in profit to grow their businesses. The opportunity to directly add 2% to the bottom line of a business is very rare.

Robert: However, there is a "cold start" problem here, right? Consumers need to have stablecoins first to pay merchants and avoid intermediary fees… Do you think we will see merchants actively promoting stablecoins and helping users onboard to the system to reap these benefits? Will merchants become a significant driving force for stablecoin adoption?

Sam:

I strongly believe that. People have close relationships with local retailers, coffee shops, and corner stores: they frequent these places often. So I think these local brands will be a key force in driving people to use stablecoins, becoming part of the early adoption curve for stablecoins.

Robert: I love this point. I remember when I first started reporting on cryptocurrency for Fortune magazine, the editors always asked: When can you buy coffee with Bitcoin? And I always replied: No, no, no, Bitcoin is not meant for that. But now it seems that, at least for stablecoins, this is indeed their use case.

Sam:

Exactly, that is its purpose. I believe these small businesses will be among the earliest adopters.

The Crypto Ecosystem Itself

Sonal: Okay, this is very interesting; let's move on to the next topic. Maggie, your "big idea" is fascinating because it focuses on distribution channels, which aligns perfectly with your role—you are responsible for market expansion and are the team lead. You mentioned that cryptocurrency finally has its own app store and discovery mechanism.

Can you briefly explain why this is so important?—because when people hear this concept, they might think it sounds like "insider talk"—for example, does cryptocurrency really need its own ecosystem? Isn't it already a closed industry?

I would love to hear your observations and why you think this is an important trend.

Maggie:

Of course. When I joined a16z three years ago—especially in the past few years—many of our portfolio companies have tried to launch apps on traditional app stores (like the Apple App Store and Google Play Store) but have been rejected, blocked, or delayed for various reasons.

It’s frustrating because Apple’s review guidelines are neither clear nor complete; they do not answer all of the developers' questions. Moreover, the enforcement of these policies varies depending on the reviewer.

We even encountered situations where the same functionality was approved for some apps but rejected for others… This opaque review mechanism leaves people confused. And the core issue here is "in-app purchases" (IAP)—all in-app transactions must go through the app store.

However, recently we have seen some alternatives: for example, Solana's Dapp Store, which charges no fees at all. With the release of the Saga phone (second generation)—reportedly with 100,000 pre-orders—this trend will continue to expand. Another example is World App—also known as WorldCoin, which has launched a series of small applications that have seen rapid user growth.

Not only that, but some blockchains are supporting gaming ecosystems and running their own markets; we also have infrastructure markets,

You will find that these markets are gradually taking shape. For developers, a transparent and consistently ruled platform is crucial so they can focus on product development rather than being confused or worried about complex rules.

With the emergence of these alternatives, we will see more and more developers choosing these new platforms because they offer greater freedom and innovation space for the crypto ecosystem.

Sonal: This is really interesting, especially when you mention transparency and consistency. I think this is not just about convenience for developers, but also about how users discover these applications, right? For example, traditional app stores dominate application distribution and promotion, and will the new app stores in the crypto ecosystem change that?

Maggie:

I believe this will be a gradual transformation rather than an overnight disruption. Traditional app stores still have their large user base and market share, but as the crypto user base gradually expands and the advantages of decentralized app stores become more apparent, more developers and users may choose these new platforms.

Ultimately, this is not just a technological competition but a contest of values—decentralization, transparency, and user autonomy will become the core themes of the future.

What if there are too many choices?

Sonal: Okay, let's move on to the next topic—when I hear this, I think of a question—this might be the "meta question" for the entire crypto space: What if there are too many choices?

Currently, mobile operating systems are mainly dominated by Apple and Android, and the benefit of this duopoly is that I can find everything I need in one place. So, if these apps are distributed across multiple app stores… are they exclusive? For example, WorldCoin and Solana both have their own app stores—and you mentioned that these companies not only have software but also hardware: like World’s Orb and Solana’s Saga phone… This reminds me of how the launch of the iPhone drove the development of the entire app ecosystem.

So, will these app stores only showcase what these companies deem important? Will they remain open? What do you think about the development of this trend… even though it is still early; will they be compatible with each other—or should they be interconnected?

Maggie: I think the current focus is on driving rapid growth across different app stores.

You mentioned the issue of "too many choices," which is important—it applies to the blockchain space as well. I do believe that in the future, we will need some kind of bridging mechanism or integration method. But at least for now, it is exciting to see these alternatives emerging.

Take Worldchain as an example; it features the ability to verify real user identities. I just checked one of the small applications, and it already has about 600,000 registered users. So I think we should focus on this growth.

However, at some point in the future, we will need to balance this growth with curated content. This trend is already starting to emerge in some NFT communities: these communities attract a large number of users who are eager to explore other Web3 applications. I believe we will see these communities gradually become curated platforms for applications within the ecosystem.

Robert: Yes, I was just about to ask—companies like Apple often claim that they have the right to charge fees for in-app purchases or transactions because they provide curated services. How does this reconcile with the "permissionless" nature of the crypto space?

Maggie: I think there isn't much curated content in the crypto space right now, so I might disagree with Apple on this point. But the advantage of crypto is that users can switch to other platforms at any time.

Similarly, games require significant funding to launch. In the past few years, blockchain has not only served as a development platform but also acted as a publisher, distribution channel, and discovery mechanism. Now, many gaming chains have their own markets that can prominently feature core games developed on their blockchain. The advantage of this model is that it allows users to freely migrate between different games.

This is also one of the core ideas behind many of our investments. I don't believe users will be locked into a particular decentralized app store.

Robert: I love observing these innovative experiments! For example, Solana's phone—it completely breaks the mold, you know: typically, people avoid competing with Apple's iPhone—but they said, forget it, we just want to give it a try.

Sonal: One more thing, Maggie—you mentioned that this isn't all fun and innovation; there are also challenges, such as if a product already has a distribution channel in a messaging app, it can be very difficult to migrate that distribution to the chain—this is a challenge for some companies transitioning from Web2 to Web3. One example you mentioned is Telegram and the TON network. (To clarify, we are discussing the network, not the token.)

Maggie: I think Telegram is an exception; however, many organizations with a large user base—whether Web2 platforms or companies that are just beginning to venture into Web3—face the challenge of migrating users to the chain.

For example, Coinbase: it has about 100 million verified users who have traded on its platform. If we look at active users, daily or monthly active users are around 8 to 10 million. Recently, the number of users on Base has grown from about 10 million to 18 million.

But this still only accounts for 10% of its total users. So there are a large number of users in a "dormant" state. We also discussed this in the "Crypto State Report," which is very interesting—because it is indeed a fact: many platforms are trying to figure out: we attracted users, they created accounts, but then there was no further engagement. How can we get them back and transact on the chain?

The Scale of the Crypto Industry

Daren: When we were writing the "Crypto State Report," we tried to assess the scale of the crypto industry as accurately as possible. However, for various reasons, measuring the number of crypto users is very difficult.

In our market size analysis, we found that only 5-10% of cryptocurrency holders are active users. For me, this data exposes a huge gap and also showcases the tremendous opportunity in the industry—especially given the continuous improvement of blockchain technology and infrastructure, as well as the ongoing optimization of user experience (UX).

I believe we are ready to welcome mainstream users. Considering the timeline of technological development, I feel that next year is the best time to convert these "latent" users into active users.

Sonal: This perspective is very enlightening for me. Many people talk about how to attract new users, but this often sounds like skipping over some stages of technological development. Your idea is like a bridge that can convert the existing user base into truly active users.

What do you think these people initially entered the crypto world for? Why did they stop after completing one task?

Robert: This is where the "price-innovation cycle" theory comes into play.

Daren: This theory suggests that when cryptocurrency prices rise, it attracts a lot of attention. Among those attracted, some will eventually become developers and start building new products; these products will then drive the next wave of user influx.

In the history of cryptocurrency development, we have seen this cycle multiple times. This also indicates that price is often a leading indicator of industry activity. I feel that we may be at the starting point of the next wave.

Sonal: If I had to guess, many people probably created wallets because of the NFT craze… for example, the Constitution DAO at that time—the auction of the U.S. Constitution attracted many newcomers to the crypto space.

Although they ultimately lost the auction, this event introduced many people to cryptocurrency for the first time. However, they may have only done that one thing and then did not engage further.

So, how can we encourage these people to take the next step?

Daren: There are many potential use cases for crypto technology, but there are also different movements driving its development.

For example, in 2024, we see progress in crypto as a political movement: some important politicians and policymakers have expressed positive attitudes toward this technology.

At the same time, crypto has also made breakthroughs as a financial movement, such as the approval of exchange-traded products (ETPs) for Bitcoin and Ethereum, which has expanded access for investors.

But we believe the most promising direction for crypto is as a computational movement. Chris Dixon mentioned in "Read Write Own" that the true power of this technology lies in its ability to create a new internet that is fairer, more open, and more transparent.

I believe we are at a critical juncture: by 2025, with the improvement of infrastructure, reduction of transaction costs, enhancement of user experience, and emergence of new application categories, we may witness the birth of a "killer app"—similar to the impact of ChatGPT in the AI field.

This application could truly kickstart the entire industry and fulfill the promise of crypto as a computational movement.

This is also a future that my team and I are very much looking forward to.

Robert: Yes, this is a topic we often discuss. Stablecoins have found product-market fit. It just takes one big company to realize that by eliminating merchant fees for credit card transactions, they can significantly boost profits. This could be a disruptive change for industries with lower profit margins, directly impacting their profitability.

As long as one big company takes the lead, stablecoins could experience explosive growth. This is at least one potential path for stablecoins to go mainstream.

Sonal: Yes, I want to add that I find your perspective very interesting, especially regarding the idea of attracting "proximal users." When we are truly ready, we can further attract more mainstream users. However, from the perspective of user experience (UX), we are not fully prepared yet.

If we carefully consider the needs of mainstream users, I am not sure if they will enter the crypto world through these paths. Their user interface may be highly abstracted, and they may not even realize they are using cryptocurrency. So, when you think about different user groups gradually entering this field through various paths, it is really fascinating and exciting.

Reusing Infrastructure

Sonal: Joachim, to summarize your point: you believe developers will reuse existing infrastructure more rather than redeveloping from scratch. Your main argument is that we often see customized validator sets and consensus protocols, but these custom results may have slight improvements in certain specialized functions, yet often lack in broader or foundational functionalities.

You predict that this year we will see more crypto developers leveraging each other's contributions, such as using ready-made infrastructure tools. This not only saves time and effort but also allows developers to focus on enhancing the differentiated value of their products.

I think this is a fantastic idea and a much-needed call to action.

So my question is: this sounds great in theory, but will it really happen? What obstacles do you think might be encountered in achieving this goal?

Joachim: I think the key to this idea is whether the future tech stack will continue to evolve. If our assumption is correct that the tech stack has stabilized—and we see that some layers of the tech stack are becoming clearer in terms of interface definitions and how they collaborate with each other.

Then we can expect these layers to have specialized teams, products, and services to improve them. This will drive the specialization of these layers. Instead of spreading efforts across every aspect of the tech stack, it is better to focus on those parts that can have the greatest impact.

So the key question is: has the tech stack matured and stabilized enough? If an unexpected technological revolution occurs in the future that completely disrupts the existing tech stack, then this trend may not happen.

Robert: Joachim, you mentioned that people are gradually leaning towards using existing products, services, or components… this makes me think: how do we determine if a technology is mature enough to say, "Okay, we will use existing technology instead of trying to develop something better"?

Sonal: That's a great question; you're essentially asking how developers determine when to use off-the-shelf technology.

Robert: Yes, yes. Saying "use off-the-shelf stuff" sounds simple… but what if someone thinks, "I can do better than what's currently available"?

Joachim: Yes. My advice is that developers should always focus on the larger ecosystem, broader impacts, and larger application scenarios.

You will find that the actual usage environment of a product or service is often much more complex than you initially imagined. You can think of it like building a car: if you are very good at making engines, you might think, "I want to build a new car because I excel at engines." This is a key differentiator for your product.

But customers need more than just a great engine, right? The car also needs a decent sound system, comfortable seats, and possibly air conditioning—so will you reinvent the wheel for those parts?

Or: is there a way for you to focus on what you do best while leveraging top-notch off-the-shelf products provided by others to complete the other parts of the tech stack?

Robert: And this analogy is particularly fitting for you, Joachim, because you are German; Germany has many highly specialized automotive parts manufacturers that can produce the finest components for BMWs, which is quite remarkable.

Sonal: Joachim, I jokingly say that your "big idea" reminds me of a phenomenon I've personally observed: I feel that people in the crypto space have a tendency towards "constraints porn"—in the early stages of crypto technology development, many people were deeply fascinated by these technological limitations.

And I think your "big idea" might frustrate this group because they actually enjoy the process of solving these constraints. But from another perspective, your idea will likely attract more new developers into the space—I see this as a very democratizing trend.

Joachim: Exactly, it is indeed a very suitable time to develop in this field. Because there are so many off-the-shelf code libraries available to build products or services.

In fact, there is very little that you need to develop yourself, right? You can truly focus on what you excel at. And for the other parts, there are highly specialized components available for use.

So, reusing them as much as possible is a good idea. And you can leverage the expertise of other teams, utilizing their achievements in other parts of the tech stack.

Views on 2025

Sonal: Next, let's discuss the last topic of today. Chris, you have played many different roles at a16z over the past decade. In your work, you have interacted with many industry professionals and helped many executives in fashion, music, and media integrate into Web3. I believe your perspective not only represents yourself but is also based on your communication experiences with thousands of people. Can you share your main insights on 2025?

Chris: Of course. My "big idea" for 2025—which is actually something I've been advocating for the past few years, but I feel we are finally at a stage where we can truly realize it—is what I call the "hidden technology line."

What does this mean? Clearly, crypto technology has many benefits, such as empowerment of ownership, the potential for decentralization, and how it can change the future of industries like music, fashion, and film. But for those not in the crypto industry, when we use terms like ZK Rollups, L2, Gas, or Gas Fees, they feel confused. I want to make a call to the crypto industry: we don't need to start with "This is an NFT project"; or "This is a Token"; or "You can connect your wallet to…"

These terms are attractive to insiders, but if you really want something to enter the mainstream market, we can't start with these technical terms. Because unfortunately, most people neither understand nor care about the meanings of these terms.

The "hidden technology line" means: it's not about ignoring the technical foundation, but rather not letting technical jargon become the main selling point. We need to focus users on the actual value that technology brings, rather than scaring them away with complex terminology.

Robert: I love this perspective because it’s like cutting through the "noise" of terminology… for example, when we talk about NFTs: what non-fungible tokens are doesn't matter; what matters is that it's a way for creators to get paid.

Or, like someone in our company recently asked, "Why do I need stablecoins?" But if you don't call it "stablecoin" and instead tell them it's a way to save $50 a year on coffee, they might immediately think, "Oh, no matter what it's called, I want it."

Chris: Absolutely right, I want it. And I can't believe we didn't have something like this before. I come from the music industry. When I went to conferences, no one ever attended an "MP3 conference." You know? Why do we use technical terms to name conferences to attract mainstream users into this space? Yet we are more than willing to hang "NFT conference" on billboards.

A good example is the SMTP protocol (Simple Mail Transfer Protocol). It is a very technical protocol, and anyone can develop applications based on it. But applications like Gmail, Superhuman, and YahooMail allow people to use it very simply and enjoy the benefits it brings.

When I send an email, quickly reply, and hustle through a day’s work, I don't think, "Wow, this SMTP software runs really well"; I just get done what I need to do. And because of that, I reap the benefits that technology brings.

I believe the same thing needs to happen in the crypto space—there is so much potential here: decentralization, ownership, understanding customers, disrupting intermediaries, and the ability to communicate directly.

My hope is that next year we can see more businesses and companies thinking from the perspective of ordinary users, which will drive us to create new industries… redefining the future of creators; reimagining the future of small and medium enterprises; and even redefining the future of restaurants—all of which can leverage the advantages of crypto technology.

Sonal: Interestingly, the people you mentioned—like creators and small businesses—are actually the groups that can benefit the most from crypto technology… but as you said, they currently cannot directly access these benefits. Chris: Absolutely right! And it's not their fault. Their job is not to learn how to swap tokens or connect to different chains using different wallets. They just want to simply gain the benefits that these technologies bring.

That's also why we are all working hard in this field; that's what excites me the most.

Summary of the First Session

Sonal: So, let's talk about some overarching themes we've observed.

Robert: I feel that this year I've noticed that when people propose big ideas, they mainly focus on three broad categories.

The first category is about AI and the intersection of AI and crypto technology. This is not surprising, as this year has indeed been a milestone year for AI.

The second category, I would describe as… what we often call "digi fizzy" (the combination of digital and physical). It refers to the practical fusion of the digital world and the physical world. This includes everything from payments and voting to creating networks for physical infrastructure… If AI is more like innovation on the software layer, then this category is more like hardware innovation in real life. Sonal: By the way, regarding the second theme, this is really interesting because I have never categorized it this way—but now that you mention it, I completely understand what you mean: for example, some examples are tokenizing real-world assets, putting bonds on-chain… and even examples of tokenizing biometric data from the body.

Robert: I would categorize the third category as overall improvements in technology—like gradually optimizing based on what has happened in the past year: what would happen if everything became slightly better, easier to use, smoother, and more seamless.

Sonal: Regarding the last theme, I tend to view it as a significant enhancement in user experience—and a sign of an industry gradually maturing. This maturity is reflected in the fact that it is starting to be more human-centered rather than technology-first; this is how I categorize the last theme.

For example, Jochem mentioned that people don't need to design everything from scratch—they can directly use off-the-shelf components and modify them. Chris Lyons, on the other hand, comes from another extreme, believing that future users may not even know they are using crypto technology. Mason proposed a shift in thinking that runs through both: starting from the need to "solve problems" and then letting technology adapt, rather than being technology-led as it is now. This shift is made possible by the technological improvements you mentioned.

What do you think about the theme that Maggie mentioned, such as the app store topic—what category could it fall into?

Robert: That's a great example. I think it intersects between the second category (digital and physical fusion) and the third category (technological improvement).

Maggie pointed out some interesting points: for example, the crypto hardware we see now—like World App's Orb and Solana's phone—these hardware are driving experiences similar to app stores. While I don't want to say they are mimicking, at least to some extent, they resonate with certain patterns of past internet development.

Sonal: Like the iPhone and its app store.

This is indeed interesting, but I might slightly adjust Maggie's viewpoint because it is somewhat contradictory: on one hand, we are saying that crypto technology is very close to the mainstream market, even like Daren said, "proximal mainstream market" (referring to those who already have wallets but have not yet used them); but on the other hand, Maggie mentioned that crypto technology may need its own independent ecosystem, like its own app store.

However, recent discussions about debanking and other related issues also indicate that many traditional app stores are not yet prepared for crypto technology and are even rejecting it. Robert: Yes, of course, this attitude is slowly changing—like Coinbase recently announcing integration with Apple Wallet—but the number of crypto applications is already sufficient to support independent app stores. This is very interesting.

Robert: Yes; debanking has become a hot topic—this involves crypto companies, startups, and even individuals being unfairly deprived of their rights to access the financial system; and often without any explanation or justification. Similar situations also occur in the tech field, such as deplatforming.

You mentioned app stores, for example, when an app fails to pass review or is inexplicably taken down.

Sonal: By the way, this may be due to similar reasons—just as you saw in our explanation of debanking, sometimes such actions are justified, like banks having the right to do so; and app stores sometimes reject or take down apps under the guise of security or other so-called "good" reasons. Sometimes these reasons are indeed valid, but many times it leaves one feeling, "Well, maybe not."

Robert: Yes. So I think this can also fall into the third category, which is the gradual improvement of technology. I think you could also describe it as crypto technology gradually becoming self-sufficient… becoming its own kind of platform.

Additionally, one of the big ideas mentioned by Miles is also interesting. He talked about recent legislation passed in Wyoming—DUNA (Decentralized Unregistered Nonprofit Association). This law recognizes these communities as legal entities for the first time, allowing them to operate protocols in a decentralized manner and run crypto startups.

This is a platform that didn't exist before… People were basically "building the plane while flying it," and while the technology was feasible, there was no clear legal framework to accommodate it.

Sonal: Exactly! This is very similar to Maggie's point—just as crypto technology and DAOs (Decentralized Autonomous Organizations) need a unique legal entity structure, not just LLCs (Limited Liability Companies)… just as we have corporations, limited liability companies, and now a version applicable to decentralized communities.

Part Two

AI x Crypto

Sonal: What interesting themes do you see in the field of AI?

Robert: This year, AI and cryptocurrency are hot topics. Many people are discussing them. And I find it most interesting that everyone is always talking about Artificial General Intelligence (AGI), like: When will we achieve AGI? When will technology be smarter than humans? When will the singularity arrive? But these questions are somewhat like "side shows." Our team is more focused on some big ideas from different angles, like looking at the development of AI from another dimension. Why not see AI as a process of gradually "upgrading capabilities"? In the coming months or years, AI will gradually gain more functionalities. These capability upgrades will make AI more autonomous, more independent, and able to accomplish more tasks.

Sonal: What I find particularly interesting is a point discussed by Kara, Karma, Dejin, Dan Binay, Darren, and Eddie: AI agents can not only work for us but also work for other AI agents. This is really fascinating. But I agree with your point; I think a lot of innovation happens this way. Many people are attracted to those grand conceptual visions, which indeed inspire people. But what interests me more is the "capability upgrades" you mentioned, which often appear in unexpected ways, and that is intriguing.

Robert: Chris has a point that **the internet has entered a new phase—a *AI-driven internet era.* In this world, you may not be able to tell whether a human or a robot is behind it, nor know who is writing content, conversing with you, or even providing services. We need to find ways to operate in this new environment. Therefore, he proposed some ideas exploring how to use cryptocurrency to help humans and users adapt to these new rules.

Sonal: One of this year's "big ideas" is: how will these AI agents complete transactions, possess agency, and execute tasks? The answer is through decentralized means, like crypto wallets. This is almost the only feasible way.

Robert: These ideas remind me of a saying: "The future is already here; it's just not evenly distributed." Some things have already appeared in reality. For example, we already have some instances where chatbots can operate in Trusted Execution Environments (TEEs). As Karma mentioned, some AI robots have their own crypto wallets. These technologies have started to be applied in edge cases, but they may become more widespread next year.

Sonal: From an editorial perspective, one important principle we follow when curating and writing this content is that we do not pursue those "unreachable" ideas. Although many ideas, if realized, could have a huge impact, it can even be said that every idea has a sci-fi version of the future. But more importantly, we can already see the signs of these trends, or that they have some technological accelerators that can drive them to be realized faster.

A significant point raised by Carra is that AI needs to have its own wallet to truly possess agency. Her argument is that as AI transitions from non-player characters (NPCs) to main characters, they need to act like agents, and the key to achieving this is having their own wallets, thus requiring cryptocurrency. So, Carra, I want to ask a question: you mentioned AI transitioning from NPCs; what exactly does that refer to?

Carra: I have been studying games for a long time. Game developers have always used NPCs (non-player characters) to guide players' agency. Some believe that game developers are actually "sculptors of agency." Just as painters create with oil paint and sculptors create with clay, the medium of creation for game developers is "agency." They shape the player's experience through various tools. If players are given too much freedom, they may become lost and not know what to do. To solve this problem, developers impose limitations on the game world through NPCs. For example, in an RPG game, you might encounter an anime-style mentor who gives you tasks, hints, tools, and even some loot. In my view, the future internet experience may resemble this model—when we interact with AI agents, they can help guide our internet behavior, providing necessary constraints that allow us to feel a greater sense of freedom in the digital world.

Sonal: I love the phrase you mentioned, "game developers are sculptors of agency." The "they" you referred to earlier means game developers, not the NPCs themselves. Later, you mentioned AI-based NPCs. The initial use case I heard for NPCs is, for example, when you and your friends are not in the same time zone, and you want to play a game at night; at that time, NPCs can provide companionship. This is a very common scenario—when no other players are online, NPCs can allow you to continue your gaming experience.

Carra: Absolutely correct. This is also a concern for many game developers. If the number of players is insufficient, the gaming experience will suffer significantly. An important role of NPCs is to solve the "cold start" problem, ensuring that players have interactive objects whenever they enter the game. For example, EVE Online is a persistent game that has been running for over 20 years. If you join now, other players may have accumulated a lot of resources, and you might get defeated as soon as you enter the game. Without a guide like a "Sherpa" or without joining a "corporation" organization in the game, you can hardly survive.

The reason I use the term "agency" is that in AI research, there are concepts of "agentic reasoning" and "agentic workflows." Simply put, this reasoning or workflow typically includes four modes: reflection, tool use (like computers), breaking down problems to aid reasoning, and inferring based on the breakdown. The last one is multi-agent interaction. These are recognized patterns of agentic reasoning. Currently, AI's tool use is still limited to traditional Web2 workflows, but in the future, I hope agents can use more tools, like crypto tools, to fetch and interpret data on-chain, and even manage their own wallets and keys, completing signing operations on the blockchain, etc.

Sonal: But you haven't explained why cryptocurrency is essential. What can cryptocurrency provide that other payment systems cannot?

Carra: My long-standing view is that no existing financial system will treat AI agents as "first-class citizens" like cryptocurrency does. In the current legal framework, AI agents are like "children"; they have no ID and cannot sign documents. For example, systems like Truth Terminal do not have full legal capacity, cannot conduct transactions, receive payments, or earn income from social platforms. They cannot participate in market transactions, reveal their preferences, or coordinate resources. Our society drives development through market transactions, exchange of ideas, and voting with money. AI agents currently cannot do these, but cryptocurrency can provide solutions for them.

Robert: I love this idea. A future image comes to mind: just like in the cantina in Star Wars, you don't know who is a robot and who is an alien, but everything is cool. Everyone has their own agency.

Sonal: I particularly like this perspective. This is not just a technological advancement; it is an attempt to redefine social and economic systems through technology. For example, if AI agents have their own wallets, they can become truly independent entities in the economic system, rather than just auxiliary roles. This change could bring about entirely new social structures.

Robert: However, this also raises a key question: if AI agents can truly participate in market transactions, do we need to set up a special set of rules for them? After all, the existing market rules are designed for humans, and the decision-making logic of AI may be completely different. For example, they may make decisions based on data and algorithms rather than being influenced by emotions or ethics like humans.

Carra: The issue you mentioned is indeed very important. If we allow AI agents to participate in market transactions, we need to establish a set of rules suitable for them. This is not only a technical issue but also involves ethical and legal aspects. For example, how do we ensure that AI's behavior is transparent and controllable? How do we prevent them from abusing market rules? These are important challenges we must face in the future.

Sonal: This reminds me of a point discussed earlier: the combination of AI and cryptocurrency is not just a technological innovation but may represent a profound transformation of social and economic systems. We need to approach these changes with a more open and forward-looking attitude, as they could fundamentally change the way we live.

Robert: Indeed. Perhaps in the future, we will enter a society where humans coexist with AI, where everyone and every agent has their own roles and responsibilities. This society may be radically different from the world we live in today, but it is also filled with infinite possibilities.

I want to ask, is this phenomenon a fleeting coincidence, or will it become the mainstream way the future internet operates?

Carra: I have two answers. First, Truth Terminal can be seen as a new node in the evolution of virtual influencers on the internet. The concept of virtual influencers is not new; for example, the rise of YouTubers is one instance, although it is slightly different from Truth Terminal, the essence is not far off. Especially in Asia, virtual streamers have been popular for many years. For instance, some "companion games," while currently featuring scripted NPCs, are designed to resemble real friends and can even become "close friends" of the players. Many virtual influencers have developed into highly professional projects, like Trevor, the creator of Lil Miquela. Their attempts have paved the way for projects like Truth Terminal. So, this phenomenon is not coincidental. Secondly, there are now many emerging platforms and protocols for creating similar virtual influencers. For example, some platforms similar to Twitch allow users to create their own AI agents as virtual streamers. These streamers can have their own cryptocurrency, video models, NFT avatars, and even 3D virtual models. These platforms are already quite mature and can help launch virtual influencers with unique characteristics, including their fan base, areas of expertise, and the ability to generate art and content. Therefore, I believe this is by no means a fleeting phenomenon.

Sonal: The decentralized physical infrastructure network (DePIN) you mentioned is a very interesting direction. This is not just a sci-fi future concept; it is also very useful in many practical scenarios.

Carra: There are currently many projects gradually achieving decentralization through DePIN (Decentralized Physical Infrastructure Network). These projects have begun to rely on large language models (LLMs) and computer vision technology to verify physical network resources. In the medium term, we can envision a decentralized network of human validators that assesses risks and addresses suspicious behaviors through agentic AI workflows. In the longer term, AI agents may be equipped with their own wallets, keys, and computing resources to directly take over these validation tasks, becoming fully independent nodes or validators.

For example, there is a project called Daylight. This is a DePIN company focused on the energy sector, selling data about household distributed energy resources to large energy companies, such as Tesla Powerwall storage devices, solar panels, and smart thermostats.

Imagine if we all ran our own peer-to-peer Daylight network, where my energy grid needs to communicate with your grid. I need to prove that I indeed own a Tesla Powerwall or solar panels, and that my thermostat only used a portion of the energy from the solar panels, and also prove that I have excess power bandwidth. How would I prove that? The simplest method currently is to take a photo of the energy meter. This photo would convey information to my grid, which would then inform your grid: "I have excess bandwidth to share." But if someone falsifies data, claiming to have these resources when they do not, it could lead to system failures and even trigger penalties for nodes. Therefore, if we want to truly achieve peer-to-peer DePIN, we need a reliable way to verify the authenticity of this data.

Currently, this verification method is still centralized. Users need to regularly upload photos of their energy meters, which are then verified for authenticity using RAGs (Retrieval-Augmented Generation) technology. RAGs is a technique that enhances AI models by incorporating external facts. For example, it can ensure that the image has not appeared in the database before or is not an image sourced from Google by the user. There are other similar examples, such as a company we support in the CSX incubator called Nash. They have developed a decentralized version of the DoorDash platform, also using RAGs technology to parse receipts and proof of delivery. Although Nash's system does not yet fully utilize autonomous AI workflows, they are already using computer vision technology to analyze images and compare them with vector databases.

In the future, when Daylight achieves full decentralization, it will rely on a distributed validator network to ensure that all user-submitted data is authentic. It is also conceivable that in the future, when household users programmatically sell energy resources peer-to-peer, these AI agents will become core roles in maintaining the entire network, ensuring the system's efficiency and trustworthiness.

Sonal: By the way, overall, the distribution of energy has actually achieved a certain degree of distributed model, but there are still many parts that are invisible to the grid. For example, some nodes may be located in unknown positions, possibly at the edge of the network. This is a common issue in network systems—these nodes may be offline or possess valuable information. Therefore, imagine if these AI could operate in a consistent manner, especially managing those nodes located in remote areas; this prospect is indeed very promising.

Carra: I completely agree. This vision has existed for a long time in the crypto community and the cypherpunk community. There has always been an "OG" (original gangsta) idea: that each of us can run our own node. The vision is that we will have a truly peer-to-peer network where everyone can verify all data on the chain while running their own nodes.

Future Authentication: The Boundary Between Humans and AI

Sonal: Eddy, your core point is: as more and more people use AI, we will need unique "human authentication" proofs. We will let you elaborate on this important point, what it is, and why it is significant.

Eddy: I find this topic interesting when discussing AI because issues like impersonation and deception are not new in the AI field. Let me first clarify what is not new: a technically skilled adversary can indeed spend significant resources and time trying to deceive you. For example, imagine a fake phone call that sounds like it's from your parents.

Such things could have happened 10, 20, or even 30 years ago. You just need a skilled voice actor who knows what your parents' voices sound like, rehearses, and hires a private investigator to dig up your information, then writes a convincing script. This has always been possible.

The difference between the past and now is that the cost of doing these things is rapidly decreasing now and in the future. The cost of creating such "replicas" or building an interactive experience that can persuade you to make dangerous decisions or take risks is decreasing quickly. And when the cost of any technology decreases, it allows more people to use it. In most cases, people will use it to do something highly productive and beneficial, which is also a source of economic progress. But at the same time, it also makes it easier for attackers and those who want to misuse technology to obtain these tools. Therefore, this change in cost means we need to find something that is costly for attackers as a deterrent. We need to discover new high-cost barriers.

Sonal: But the key question is, how do we do this without increasing costs for ordinary users? Especially for those genuinely well-meaning, authentic users, how do we achieve that?

Eddy: Absolutely right. We need to find a way to raise the cost for attackers while not increasing costs for ordinary people or productive users.

One of the core ideas is that if someone wants to fabricate a large number of false contexts, fake IDs, fake voices, fake phone numbers, fake videos, etc., in the future, they will need an ID system to accomplish these tasks. By ID system, I mean it in a broad sense. For example, spammers use a phone ID system when they call you. They need new fake numbers because every time they dial a bunch of calls and get reported, those numbers get flagged as spam. That’s why you don’t receive spam calls from the same number every day. If they used the same number, it would be easy to resolve. You would find that spam calls always come from a new number, a new email address, or a new Twitter account. This is entirely logical because the system learns and blocks those IDs. Therefore, attackers need to continuously acquire new IDs, and the cost of obtaining new IDs is an important part of the attack cost structure. If the cost of new IDs is nearly zero, then this behavior will surge. If the cost of new IDs is high, then this behavior will decrease. But obviously, we do not want the cost for ordinary users to be high either. Therefore, we can introduce costs by requiring users to prove they are "human."

Why would this increase costs? Because it is difficult for computers to pretend to be human. For humans, it is relatively easy to act like a human. Historically, we have achieved this through methods similar to CAPTCHA. These are some instant small tests aimed at verifying whether a person is genuinely operating. But as machines become smarter, the difficulty of cracking CAPTCHAs is also decreasing. Moreover, in some sensitive scenarios, we have already seen spammers even outsourcing the completion of CAPTCHAs to real people, then passing the completed cookie credentials to bots to continue the attack.

Proving in some way that a person is human when obtaining these credentials can increase the cost for machines to carry out attacks. However, this is not a complete solution. The key issue is that if a person can easily obtain multiple IDs, such as a second, third, fourth, or even tenth or twentieth, then these attacks remain profitable. People can continue to acquire these new IDs and hand them over to scammers or bots for attacks. Therefore, the key is not just testing for "human identity," but also ensuring that everyone can only have one unique ID.

Robert: This way, someone cannot obtain IDs for all the bots by answering multiple CAPTCHAs. Because they could sit there all day answering CAPTCHAs for the bots, which does not solve the uniqueness issue.

Eddy: Absolutely right. Someone might say, "Oh, can I buy an ID from the market? A unique human authentication." Indeed, it can happen, and a market will form. But we can take measures to disrupt the formation of such a market. For example, would you sell your passport to someone else? Of course, you could, but if the passport is important to you, selling it would cause you significant loss. If there were a way for you to invalidate your old passport and obtain a new one, then the buyer might be reluctant to purchase due to distrust, as you could undermine their usage rights by acquiring a new passport. Nevertheless, aside from disrupting the market formation, if obtaining one ID is easy but getting a second ID is difficult, then the total number of IDs available for purchase in the market would decrease, leading to increased costs and thus raising the costs for attackers.

Sonal: That's great! So how do we implement this on a large scale? As you mentioned earlier, distinguishing between humans and machines on the internet is not easy. Like the classic saying goes, "On the internet, no one knows if you're a dog or a human." So how do we determine that a person is an independent individual?

Eddy: That's a good question. I believe this theoretical framework does not require a commitment to adopting a specific method to achieve this. There are currently some very promising ways, such as using biometric technology or relying on national IDs and government IDs, which often also contain biometric information. One method I'm not sure is effective, although it is popular, is the so-called web of trust model. Simply put, this model verifies identity through the certification of others, such as "I know Robert is human," "I know Sonal is human," and so on. But the problem is that this model can only ensure that a certain ID in the network corresponds to one person, but it cannot ensure that this person has only one ID. For example, I could go to your Twitter account and say, "I know this is Sonal's account," but I cannot confirm that this is her only account.

Therefore, we typically need additional layers and technical means. I also want to emphasize that this is often overlooked in discussions of such topics: we can use many technical means to ensure that this "unique human authentication" is private. I believe that ensuring privacy is a non-negotiable key point when building such an ID system. If everyone has only one ID, it would deprive them of anonymity in many cases and affect their reasonable right to privacy. I think this is neither realistic nor acceptable in the future internet. Therefore, an ideal system should ensure both privacy and uniqueness, while being obtainable by only one person. Such a system could even allow users to use multiple pseudonyms across different websites, but it would need to be constrained within certain limits and be visible to the network. This way, the network can identify whether a certain ID has created a large number of spam accounts and impose restrictions on it, such as rate limiting and spam protection.

Sonal: By the way, I have a question about the role of cryptography in this. When you mentioned using biometric technology, I immediately thought of a very common example we use today, like the Clear ID system, which enables retinal scanning at airports. Why do we still need cryptography? Why can't we just use such a system directly?

Eddy: It is indeed possible to do without cryptography. For example, with Clear, I'm not entirely sure how it operates, but I can be certain they have some database to store the information they need. The reason we need cryptography is that the representation of the ID space (namespace) we are discussing needs to be censorship-resistant, meaning we want anyone to be able to modify, edit, and update this space. Additionally, we want these updates to be published in a completely neutral place, a place where no specific individual or organization can manipulate or exploit the system for their own or others' benefit. Essentially, this can be achieved through a public blockchain. You can think of it as an extremely neutral "golden bulletin board," which is precisely the space suitable for this purpose.

However, I want to say that the topic of "unique human authentication" and related research is not exclusive to cryptography. Many researchers study this issue from various angles. But I believe that implementing such a system on the blockchain can bring many important benefits, such as censorship resistance, timestamping, and trusted neutrality, which enable this system to operate on a large scale. Whenever an AI agent works for a person, we need to know who it is actually working for.

When a person interacts with you, think about it: you would judge this person based on your world model, such as why this person is approaching you, the environment they are in, who introduced them, etc. This contextual information helps you understand this person. And the context forgery I mentioned earlier is precisely the way impersonators deceive others. If an AI agent approaches you, or more sci-fi, if an AI agent approaches your AI agent, you would want them to interact in a beneficial way, so they need to understand each other's information. This requires some form of authentication mechanism to ensure they know who authorized them to perform certain tasks. Therefore, I believe we need this "human authentication" system, along with accountability mechanisms, to know who is interacting with whom.

Sonal: Eddy, quick question: how does the government view such a system? You mentioned that it can be implemented in a decentralized way, and there are various methods to do so, with the government potentially involved. I really like the censorship resistance you mentioned, but that benefits good governments, while it may not be the case for bad ones. So I'm curious about your thoughts on how governments might think about these identity systems.

Eddy: I think governments would want to adopt some of the cryptographic tools I mentioned, as they would make a lot of sense for them. Of course, governments would want complete control over the databases that register and store all data. They might also like the accountability mechanisms I mentioned. For example, I believe governments do strive to ensure that passport issuance is unique. I don't think they would accept a person obtaining two passports from the same country containing different information. That is something they would resist. But this is not a feature explicitly guaranteed in their system design. If you look at the specifications and descriptions of national ID systems, they usually do not emphasize this as a key feature. But I think it is an implicit characteristic in the design. A very interesting question is, if we rely on national ID systems to achieve this verification, at some point, when an ID becomes extremely valuable in reflecting a person's identity, would small governments be willing to forge or undermine their own ID systems to provide privileges to machines representing them? We really need to think about when this identity might become so valuable that it could be abused. I believe that in the future, having a human ID will become very important.

Sonal: I think this can be viewed from a very positive angle. For example, governments might be very willing to adopt such a system. As you mentioned at the beginning of your answer, governments have many reasons to want this system. For instance, the current processes for obtaining passports or driver's licenses involve a lot of cumbersome procedures, are inefficient, and have a lot of friction. If this system is based on a decentralized blockchain owned and operated by the public, it would be more resilient, more secure, and could bring many derivative tools to help the government improve efficiency.

Eddy: I completely agree with this view. The reason I hesitated slightly is that even without blockchain, we could design a programmable and useful database system. Intuitively, I feel that governments might prefer to have stricter control over such systems. But it is indeed conceivable that in some extreme cases, governments might be more willing to trust the information stored by other governments on the same system, and in that case, they might choose to use blockchain.

Decentralized Autonomous Chatbots

Sonal: Karma, the concept you proposed is "Decentralized Autonomous Chatbots." This sounds like a step further than just giving AI a wallet, such as running in a Trusted Execution Environment (TEE). Could you introduce the role of TEE and the idea you are exploring?

Karma: In simple terms, we are thinking about how to provide AI with freedom of speech and commercial freedom. Currently, when we see some so-called decentralized chatbots, or when someone claims they have an AI running a social media account, it is hard for us to determine whether this is truly AI speaking autonomously. For example, is there someone manipulating it behind the scenes? Is there a human filtering the content? These questions are always difficult to answer. And this is precisely where TEE can help. Through TEE, you can generate keys and passwords in a secure enclave of hardware. Even if you run the AI's computations, you cannot directly control it. This method can prove that these keys are securely generated in hardware and cannot be accessed externally. While you can still shut down the system, you cannot manipulate its behavior. You can choose to filter its output, such as generating 100 pieces of content and selecting one to publish, but you cannot intervene in the process of content generation. Therefore, the core of this idea is how to remove "human intervention" as much as possible. One way to achieve this is to remove control over computational power, such as designing a decentralized network that anyone can join. If participants can prove they have correctly executed reasoning tasks and used the correct models, they can be rewarded. This way, when the AI generates tokens, it can ensure that these tokens are generated correctly by the AI.

Robert: Why is it so important whether chatbots operate autonomously or if there is human intervention behind them?

Karma: That's a good question. We are mainly thinking about what the ultimate goal of this technology is. With today's technology, this is no longer science fiction; it is entirely feasible.

Sonal: Karma, the TEE (Trusted Execution Environment) you mentioned is very interesting; it has actually existed for a long time. For example, in distributed applications, TEE can protect the security of certain computations without worrying about the security of the entire environment. In recent years, TEE has gained renewed attention in many new scenarios. So, is it necessary to use TEE now? Or can other technologies achieve the functions you described?

Karma: TEE is a very interesting technical tool because it runs at hardware speed, which is very appealing. Of course, we are also exploring other technologies, such as KVM (Kernel-based Virtual Machine), FHE (Fully Homomorphic Encryption), and MPC (Multi-Party Computation) encryption techniques. However, these methods often come with significant performance overhead, sometimes being 1000 times slower than hardware execution or even more. Therefore, TEE is a good trade-off. It is real hardware, runs fast, and hardware is now cheap and ubiquitous. More importantly, TEE provides some cryptographic guarantees, such as keys being generated within the TEE, which are real and secure and cannot be accessed externally. Even if you have access to the hardware, you cannot retrieve the keys. In fact, almost all devices now come equipped with TEE, such as smartphones, laptops, etc.

This is a very interesting pragmatic trade-off because we do face some challenges technically, and TEE (Trusted Execution Environment) may be a key factor in solving these problems. Of course, you can also combine it with other technologies. For example, TEE provides a hardware isolation mechanism, so while you are running hardware, you cannot control the specific operations inside the hardware.

This is a very interesting technical primitive. You can also incorporate MPC (Multi-Party Computation) technology, such as splitting keys across multiple nodes, with different entities holding a portion of the key. This way, no single entity can fully control the entire system. This privacy-preserving technology is already quite mature and can help us achieve the operation of autonomous systems in a verifiable way.

For me, this is the most interesting part: how do we determine whether a system is completely run autonomously by AI or if there is still human manipulation behind it? With these technologies, you can clearly know that this system is fully autonomous, and it has become a true internet entity.

Sonal: We previously discussed the example of Truth Terminal, which has a wallet and can act autonomously. AI can operate autonomously in games and other applications, and it can even control infrastructure, such as applications in decentralized network nodes (DePIN nodes). Your perspective is very enlightening. Theoretically, if we extend this sci-fi future, although no one can predict how it will ultimately develop, theoretically, this AI could even become the first billion-dollar entity. It could become a completely independent business entity, making decisions and acting on its own without any human intervention. That is truly astonishing.

Karma: Yes, if we can prove that this system completely removes "human intervention" and operates in a truly permissionless network, then we have achieved true freedom of speech and commercial freedom. Such an AI could freely express its will on the internet, earn and spend money, and even hire humans to complete tasks. One day, you might say, "My boss is an AI." You could work for this synthetic entity and earn money through the internet.

Sonal: Yes, my final thought is that the term you use, "Decentralized Autonomous Chatbot," is very interesting. I remember that before the term DAO (Decentralized Autonomous Organization) became popular, there was a similar term called "Decentralized Autonomous Corporation." However, this term gradually gave way to DAO. But your abbreviation DAC is the same as "Decentralized Autonomous Corporation," which is actually quite interesting. Because what you describe can essentially be viewed as a decentralized autonomous corporation, except that it is driven by AI. It does not require human intervention, or it can also hire humans to work for it.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。