Author: YBB Capital Researcher Zeke

Introduction

Starting from the inscription craze and ending with the election of the first crypto president, 2024 is drawing to a close. This year, the crypto space experienced an unusual “bull market,” with altcoins underperforming and meme tokens gaining dominance, but ultimately, everything has converged back to BTC. While there were some low points and lingering frustration, overall, the crypto industry has been moving in a more positive direction. Looking ahead to 2025, there are many areas worth paying attention to. This article will provide a brief outlook for the coming year based on recent viewpoints.

I. About AI

At the current stage, blockchain projects often overcomplicate their technological implementations in pursuit of conceptual perfection, which ultimately affects user interaction and experience. Projects built around the Intent architecture tend to be even more complex. Whether centralized (e.g., TG Bot), structured (combining on-chain and off-chain pre-processing), or distributed (e.g., Solver + Executor architectures), these intent-based projects often share some common issues. For example, users still need a certain level of understanding of DeFi, and the expression of intent must be clear, accurate, and simple. When users present complex and vague intentions, current intent-based projects often struggle to deliver, with limited scope for execution. Since Paradigm introduced this concept in mid-2023, intent-focused projects have largely been all talk and little action, offering little help in guiding new users or lowering the barriers to entry. However, it’s clear that, based on the development path of Ethereum Layer 2, the demand for such solutions is urgent.

Looking back at the development of Layer 2 in recent months, leading projects like OP Superchain have increasingly expanded their alliances. ZkSync’s Elastic Chain and Arbitrum Orbit are also following this path to form their own alliances. These alliances will enable internal interoperability through clustered solutions to alleviate the fragmentation and lack of interoperability in the Ethereum Layer 2 ecosystem. In the future, competition among dozens of chains will narrow down to a multi-party contest. However, from a broader perspective, with the crypto market improving, new Layer 2 projects like Movement and Fuel are launching their mainnets to capture scarce liquidity in the altcoin market. For projects beyond the first tier, fragmentation and lack of interoperability are still worsening. Virtual machines based on different architectures may not even have wallet plugins that can communicate with each other. For ordinary blockchain users, the entire Layer 2 ecosystem is increasingly complex, which will create significant barriers for the development of non-financial applications.

For Ethereum to onboard new users, ecosystem alignment is a key prerequisite. An ecosystem that requires users to be semi-geeks will never achieve “mass adoption.” Looking at the reverse trend of Solana and Ton this year, strategies that lower user thresholds and offer more Web2-like experiences have clearly played a vital role in ecosystem growth. To put it directly, these two ecosystems have only reduced the difficulty of asset issuance and made the use of the blockchain almost seamless for users. Thus, Ethereum must adopt an integrated approach focused on user experience. However, given the core developers’ consistent open stance, it is unlikely that the entire Layer 2 ecosystem will be forced to align.

I believe that the solution to this problem lies in AI browser agents. Early on, many people envisioned that AI would revolutionize app interactions, transitioning from single-point to cross-app operations, creating a super app. To illustrate, in the case of travel, once the AI receives a user’s travel request, it can automatically handle ticket bookings, custom travel routes, dining, and time schedules. If the AI also has long-term memory, it could arrange future trips tailored to the user’s needs.

Now, Google is about to launch an AI browser agent powered by Gemini, called Project Mariner. In the demo presented by Jaclyn Konzelmann, Director of Google Labs, when a user installs the AI agent extension in Chrome, a chat window pops up on the right side of the browser. The user can instruct the agent to perform tasks like “Create a shopping cart from this list at a grocery store.” The AI agent then automatically navigates to a grocery platform, adds the items to the cart, and moves to the checkout page. Once everything is confirmed, the user completes the purchase (the agent does not have payment authority). A similar product from OpenAI will also be launched next month.

It is worth noting that while Google’s Project Mariner is currently only available to select testers, I have already experienced similar agents developed for general users by certain crypto projects. From a few hours of trial, the agent’s accuracy in executing complex or vague tasks is around 60–70%, and it can autonomously perform tasks like token trading on decentralized exchanges across multiple blockchains, including cross-asset transfers to Layer 2 solutions. In this process, I only need to provide the agent with the intent and input my wallet password.

Of course, this platform still requires the use of centralized model APIs. So, what’s the intersection between Crypto and this development? I believe that AI browser agents will not only improve the user experience for intent resolution but will also drive the growth of AI wallets, decentralized computing, and decentralized data projects in the coming year.

Think about a simple question: Why has the beautiful vision of an AI agent only become a reality now, after years of rapid AI development? Looking back at the development of OpenAI, it is clear that language models have evolved faster than image generation models because the internet itself is a vast corpus that provides endless text material for training. What has held back the development of language models is more about computing power and energy constraints. On the other hand, agents require a lot of manual tagging and feedback, and the reasoning process is costly. Crypto is naturally suited to incentivize labor acquisition. In this economic system, upper-level users can provide large amounts of tagged data and feedback in a decentralized way to earn tokens, while the lower level can integrate decentralized computing and data projects. Once training is completed, it can be integrated with wallets and DeFi projects through SDKs, leading to a true AI wallet and ultimately creating a closed-loop system. Other AI agent ideas can be derived from this model, as any AI agent designed for Web3 will need computing power, tagging, and feedback to “grow.”

2. Stablecoins

Stablecoins have always been a crucial battleground and remain a high-barrier sector within Crypto. As for their application value, they have garnered widespread recognition even outside the industry. For instance, this year, several major players in traditional finance entered the stablecoin market, including PayPal with its PYUSD, BlackRock’s collaboration with Ethena on USDb, and VanEck’s introduction of AUSD, which serves regions such as Argentina and Southeast Asia.

As Tether and Circle continue to strengthen their dominance in this sector, the new entrants into the stablecoin space have gradually diverged into two categories. First, issuers of fiat-backed stablecoins have begun to focus on emerging markets in South America and specific application scenarios. Meanwhile, algorithmic stablecoins are increasingly being designed with low-risk financial products as their underlying assets, such as Ethena and Usual, which we mentioned in our previous article. Looking at trends, it seems that next year, more delta-neutral stablecoins will compete for short-selling liquidity in CEX, while hedging assets will gradually expand from BTC and ETH to riskier, less liquid public chain tokens, aiming to capture the last remaining niche markets. As for Usual-like stablecoins backed by mid-to-short-term U.S. Treasury bonds, I believe the focus will still be on innovations in protocol tokens and yield methods, as there are no better choices for underlying assets than mid-to-short-term government bonds. However, compared to the limited liquidity in CEXs, this type of stablecoin faces less competition and has a greater upper limit of potential.

Overall, the development of stablecoins is gradually moving toward more stable underlying assets and decentralized governance. However, what I hope for the most is the emergence of fully decentralized and non-overcollateralized stablecoin protocols in the coming year.

3. Payments

With the regulatory approval and accelerated adoption of stablecoins in various countries, the payment sector downstream of stablecoins will also become a new competition focus. Heterogeneous public chains like Solana and Move, which offer high TPS and low gas fees, will serve as the main infrastructure for payment applications. Traditional payments are already a mature and highly competitive red ocean market. So, what can blockchain bring to the table in terms of innovation? There are two points that are frequently discussed. First, blockchain can optimize cross-border payments, eliminate pre-funding requirements, and make cross-border remittances faster, cheaper, and easier, thus solving the problem of trillions of dollars tied up in pre-funding within traditional systems. Second, it serves emerging markets. As mentioned in my previous articles, the application value of stablecoins has already been demonstrated in regions such as Asia, Africa, and Latin America. The strong financial inclusion offered by stablecoins allows residents of third-world countries to effectively deal with the high inflation caused by unstable governments and enables them to participate in global financial activities and subscribe to the world’s most cutting-edge virtual services.

The concept of “PayFi,” introduced by Solana Foundation manager Lily Liu at the 7th EthCC conference, offers more possibilities for combining blockchain with payments. This concept involves two core ideas: first, real-time settlement, or T+0 settlement. PayFi enables same-day settlement or even multiple settlements within a day, eliminating delays and complexities inherent in traditional financial systems, significantly increasing the speed of capital flow. Second, Buy Now, Pay Never (BNPL), for example, where a user deposits $50 into a lending product to buy a $5 cup of coffee. Once the accumulated interest reaches $5, it is used to pay for the coffee, and the funds are unlocked and returned to the user’s account.

There are many ideas that can stem from this. For example, in terms of use cases, financing needs for emerging projects can be met through PayFi on the blockchain, providing a safer and more transparent means of entry and exit. Currency exchanges during travel will no longer require various physical financial institutions. Payment and receipt times will be more flexible (delayed receipt of funds to earn interest, early payment to get discounts). Moreover, the methods of generating revenue will diversify. Besides earning interest by depositing stablecoins into lending products, I personally believe that the variety of stablecoins should also allow for easy conversion. In the future, as more emerging stablecoins flood the market, users will be able to choose the most suitable type of stablecoin based on their risk tolerance, while simultaneously receiving stablecoin protocol tokens and higher interest on stablecoins. If this payment system becomes mainstream, its growth potential in DeFi will be enormous.

4. DEX

As mentioned in the first section, the fragmentation of Layer 2 solutions and their lack of interoperability pose significant challenges. However, there is an additional issue with this development path: excessive block space. The development of Infra is far outpacing the growth of DApps. This problem is likely to result in the natural elimination of many long-tail chains within a few years, and it is also a significant headache for Ethereum, which struggles to receive positive feedback from Layer 2 solutions due to its flawed DA (Data Availability) pricing mechanism.

Looking back, the public chains that have experienced this countercyclical growth have mainly relied on their strong communities, ecosystems, and marketing advantages. These advantages have been provided to asset issuance platforms, enabling rapid growth in overall TVL. Therefore, not every Layer 2 solution can replicate this “attention economy” model. The lack of killer applications will continue to be a real issue in the coming year. Continuing with the trends, one possible avenue is the emerging demand for AI agents, which we mentioned earlier. In the short term, other clear trends include on-chain order book DEXs, privacy solutions, payment-related stacks, and decision-making tools.

Personally, I am quite optimistic about on-chain order book DEXs becoming the mainstream of the next generation of decentralized exchanges. After all, looking at the development of AMM, the complexity of its technological path is continually increasing, but the incremental efficiency gains are becoming less and less significant. We have also discussed this in our previous articles about Uniswap. However, for Layer 2 solutions, the limitations in performance and gas costs are still very obvious. Therefore, innovations in matching algorithms and gas solutions will be key challenges.

5. Asset Issuance Remains the Mainstream

From 2023 to today, from inscriptions to the current AI Meme platforms, the method of asset issuance has been a hot topic throughout the past year. If we broaden the time frame, asset issuance has been the core theme of the crypto space ever since the ICO era. The only thing that has changed is the external packaging and the barriers to entry.

On the positive side, the demand for user participation in the market has driven the early development of Infra and DeFi. As this technology has gained recognition and acceptance, blockchain has entered the mainstream and integrated with the real world. On the negative side, the game has become more purely competitive and absurd. The lowered difficulty of asset issuance also means that this dark forest has become more dangerous. Nowadays, all it takes is a few clicks, some images, and a few lines of text, and you can start a large-scale zero-sum game. But why not redirect this to a more positive side and use it to promote progress within the industry?

For example, some of the current AI Meme projects are starting to evolve into practical agents, rather than the earlier versions of AI agents that spouted nonsense. The recently popular DeSci could be seen as a “scientific version of ICOs.” Although its core is driven by memes at the moment, in the long run, combining blockchain’s advantages could make traditional scientific research more transparent, easier to disseminate, easier to fund, and easier to collaborate. However, whether it will succeed in practice, and how it will evolve, is still uncertain.

In fact, the idea behind DeSci is similar to what I mentioned in my article on GameFi, such as how blockchain can effectively promote the development of independent games amidst the financial and personnel shortages faced by small studios. The issue with blockchain-based fundraising is that the barriers to asset issuance are too low, there are too few restrictions, and the fundraising capacity is too strong (which could also be attributed to the extremely low entry barriers on the blockchain). How to regulate the use of funds through rules, forcing project teams to continuously create genuinely valuable products, is something we should focus on.

Let the players compete, and let the builders advance. This is the premise for the continuous development of blockchain. In the coming year, we may witness more iterations of the “ICO” model, but what I hope for is that, in this grand game, we will see the emergence of the next “DeFi Summer.”