Author: Zeke Source: YBB Capital Translation: Shan Ouba, Golden Finance

The crypto world in 2025: Breaking the Deadlock and Prospecting for Reconstruction

From the craze of inscriptions to the election of the first "encrypted president", 2024 is about to end. This year, the crypto market experienced an unusual "bull run", with altcoins underperforming, meme coins taking over, and ultimately everything coming back to Bitcoin. Despite the lows and continued disappointments along the way, overall the crypto industry is moving in a more positive direction. Looking ahead to 2025, there are many areas to watch. This article will provide a brief outlook for the coming year based on a near-term perspective.

Currently, blockchain projects tend to pursue conceptual perfection and overly complicate technical implementation, affecting user interaction and experience. Projects built around intent architecture are particularly complex. Whether it is centralized (such as TG Bot), structured (preprocessing that combines on-chain and off-chain) or distributed (such as Solver + Executor architecture), these intent-based projects generally have some common problems. For example, users still need to have a certain degree of DeFi understanding, and the expression of intent must be clear, accurate, and concise. Current intent-driven projects often struggle to respond and have limited execution scope when users present complex or ambiguous intentions. Since Paradigm proposed this concept in mid-2023, intent projects have mostly remained at a theoretical level and have done little to guide new users or lower the entry barrier. However, judging from the development path of Ethereum Layer 2, the need for such solutions is very urgent.

Looking back at the development of Layer 2 in the past few months, leading projects such as OP Superchain are continuing to Expanding the alliance, ZkSync’s Elastic Chain and Arbitrum Orbit are also following suit, trying to form their own alliances. These alliances will achieve internal interoperability through aggregation schemes to alleviate the fragmentation and lack of interoperability of the Ethereum Layer 2 ecosystem. In the future, this multi-chain competition will gradually shrink to a contest between a few parties. However, from a broader perspective, as the crypto market picks up, emerging Layer 2 projects such as Movement and Fuel are launching mainnets in an attempt to compete for scarce liquidity in the altcoin market. For non-frontline projects, fragmentation and interoperability issues continue to worsen. Virtualization of different architecturesMachines may not even have wallet plugins capable of communicating with each other. For ordinary blockchain users, the Layer 2 ecosystem has become increasingly complex, which will become a major obstacle to the development of non-financial applications.

For Ethereum to welcome more new users, the integration of the ecosystem is a key prerequisite. An ecosystem that requires users to have a geeky level will never achieve "widespread adoption." Looking back at the growth of Solana and Ton this year against the trend, the strategy of lowering user thresholds and providing more Web2-style experiences has obviously played an important role in ecological growth. In other words, these two ecosystems only simplify the difficulty of asset issuance, making users' operations on the blockchain almost imperceptible. Therefore, Ethereum must adopt an integration strategy centered on user experience. However, considering the consistent openness of core developers, it is unlikely that the entire Layer 2 ecosystem will be forcibly integrated.

I think AI browser agents are the answer to this problem. In the early days, many people imagined that AI would completely change the way applications interact, transitioning from single-point operations to cross-application operations, creating super apps. For example, in a travel scenario, once AI receives the user’s travel needs, it can automatically complete ticket booking, customized travel routes, meal arrangements, and schedule planning. If AI has long-term memory, future trips can also be automatically arranged according to user needs.

Currently, Google is about to launch an AI browser agent called Project Mariner, based on the Gemini model. According to a demonstration presented by Google Labs Director Jaclyn Konzelmann, when a user installs the AI ​​agent plug-in in the Chrome browser, a chat window will pop up on the right side of the browser. The user can instruct the agent to perform a task, such as: “Create a shopping cart based on this list at the grocery store.” The AI ​​agent then automatically navigates to the grocery platform, adds the item to the cart, and jumps to the checkout page. The purchase is completed after the user confirms (the agent does not have payment authority). OpenAI will launch a similar product next month.

It is worth noting that although Google’s Project Mariner is currently only available to a small number of test users, I have experienced AI agents developed by some crypto projects for ordinary users. . Over the course of several hours of trialling, the agent was approximately 60-70% accurate at performing complex or ambiguous tasks. It can autonomously trade tokens on decentralized exchanges (DEX) on multiple blockchains and even move across assets to Layer 2 solutions. In this process, I only need to provide the agent with operation instructionsPicture and enter the wallet password.

Of course, the platform still needs to call the centralized model API. So, where does cryptocurrency intersect with this development? I believe that AI browser agents will not only improve the user experience of intent resolution, but also promote the development of AI wallets, decentralized computing, and data projects.

Think about a simple question: Why has the wonderful vision of AI agents not been realized until now?

Looking back at the development of OpenAI, the reason why language models (LLM) develop faster than image generation models is because the Internet itself is a huge text corpus, which provides training A wealth of material is provided. What limits the development of language models is more computing power and energy bottlenecks. AI agents require a lot of manual annotation and feedback, and their reasoning process is costly.

Cryptocurrencies are naturally suited to incentivizing labor acquisition. In this economic system, advanced users can earn tokens by providing large amounts of annotated data and feedback in a decentralized manner, while the bottom layer can integrate decentralized computing and data projects. Once the training is completed, it can be integrated into wallets and DeFi projects through SDK to create a true AI wallet and eventually form a closed loop. Ideas for other AI agents can also be derived from this model, as any AI agent targeting Web3 will need computing power, annotation, and feedback to grow.

Stablecoins have always been an important battlefield in the encryption market, and it is also an area with extremely high entry barriers. Its application value is not only widely recognized in the industry, but also attracts attention in the traditional financial field. For example, this year PayPal launched PYUSD, BlackRock cooperated with Ethena to issue USDb, and VanEck launched AUSD for Argentina and Southeast Asia, etc., which all indicate that traditional financial giants are accelerating their deployment in the stablecoin market.

As Tether and Circle continue to consolidate their dominance in the industry, new players entering the stablecoin market have gradually divided into two categories:

1. Issuers of legal currency-backed stablecoins have begun to pay attention to emerging markets such as South America and specific application scenarios;

2. Algorithmic stablecoins increasingly use low-risk financial products as underlying assets, such as Ethena and Usual (we mentioned them in previous articles).

Judging from the trend, more "delta-neutral" stablecoins will flood into the market next year to compete for short liquidity on CEX. The scope of hedging assets will gradually expand from BTC and ETH to public chain tokens with higher risks and less liquidity, striving to occupy the last niche area of ​​the market.

As for stablecoins like Usual that are backed by short- and medium-term U.S. debt, I think the focus will still be on protocol tokens and revenue innovation, because in In terms of underlying assets, short- and medium-term bonds are still the best choice. In addition, compared with the limited liquidity in CEX, this type of stablecoin has greater potential and room for development in areas with less competitive pressure.

Overall, the development of stablecoins is gradually moving towards more stable underlying assets and decentralized governance. But what I’m most looking forward to is seeing the emergence of fully decentralized stablecoin protocols next year that don’t require over-collateralization.

As stablecoin supervision in various countries gradually comes into effect and accelerates its popularity, the payment field will become a new focus of downstream competition for stablecoins. Heterogeneous public chains like Solana and Move, with their advantages of high TPS and low gas fees, will become the main infrastructure for payment applications.

The traditional payment market is quite mature and highly competitive. It is a typical "Red Ocean market". So, what innovations can blockchain bring to the payment field? There are usually two directions:

1. Optimize cross-border payments

Blockchain can Eliminate the need for pre-financing, make cross-border remittances faster, cheaper and more convenient, and solve the problem of trillions of dollars being locked in pre-financing in the traditional system.

2. Serving emerging markets

As I mentioned in my previous article, stability The application value of the currency has been verified in Asia, Africa and Latin America. The strong financial inclusion of stablecoins enables residents of the third world to effectively cope with the high inflation caused by instability, participate in global financial activities, and enjoy cutting-edge virtual services.

At the 7th EthCC Conference, Solana Foundation Manager Lily LiuThe concept of **"PayFi"** was proposed, bringing more possibilities to blockchain payments. PayFi contains two core concepts:

1. Real-time settlement (T+0 settlement)

PayFi can achieve same-day or multiple settlements, eliminating the delays and complexities inherent in the traditional financial system, and greatly increasing the speed of capital flow.

2. "Buy Now, Pay Never"

For example , the user deposits $50 into a lending product and buys a $5 cup of coffee. Once the accumulated interest reaches $5, the interest is automatically paid for the coffee, and the remaining funds are unlocked and returned to the user's account.

Many ideas can be derived from this. For example, in terms of use cases, the financing needs of emerging projects can be met through PayFi on the blockchain, providing a safer and more transparent way to enter and exit. Currency exchange while traveling will eliminate the need for various physical financial institutions. Payment and collection times will be more flexible (delay collections to earn interest, early payments to receive discounts). In addition, income generation methods will also be diversified. In addition to depositing stablecoins into lending products to earn interest, I personally believe that the types of stablecoins should also be easily exchangeable.

In the future, as more and more emerging stablecoins enter the market, users can choose the type of stablecoin that best suits them based on their risk preferences while earning protocol tokens. coins and higher stablecoin interest. If this payment system becomes mainstream, its growth potential in DeFi will be huge.

As mentioned above, the fragmentation and lack of interoperability of Layer 2 are becoming important factors for the development of the encryption market. hinder. However, there is another problem with this development path: excess block space. The development speed of infrastructure (Infra) far exceeds the growth speed of DApp (decentralized applications).

This imbalance may naturally eliminate many long-tail chains in the next few years, which is also a headache for Ethereum. Due to the imperfect data availability (DA) pricing mechanism of Ethereum, the positive feedback that Layer 2 brings to the Ethereum mainnet is very limited.

Looking back on the past, those public chains that grew against the trend in the bear market basically relied on strongLarge community, complete ecosystem, and marketing advantages. These advantages are mainly reflected in the asset issuance platform, driving the rapid growth of total lock-in volume (TVL). Therefore, not all Layer 2 can replicate this "attention economy" model. The lack of hit apps will continue to exist in the coming year.

In terms of future trends, one possible development direction is the AI ​​agent demand we mentioned before. In the short term, other clear trends include:

•On-chain order book DEX

•Privacy solutions

•Payment-related technology stack

•On-chain decision-making tools

I am personally optimistic about the development of on-chain order book DEX and believe that it will become the mainstream of the next generation of decentralized exchanges.

After all, judging from the development of AMM, the complexity of its technical path continues to increase , but the marginal efficiency improvement is getting smaller and smaller. We have discussed this issue in previous articles about Uniswap.

However, for Layer 2, the limitations of performance and gas costs are still very obvious. Therefore, innovation in matching algorithms and gas fee optimization will become key challenges.

From 2023 to the present, from Inscription to AI Meme platform, the asset issuance method has been used throughout the past year. If we extend the timeline, asset issuance has been the core theme of the crypto market since the ICO era. The only thing that has changed is the exterior packaging and entry threshold.

The positive side is that the market's demand for user participation has promoted the development of early infrastructure and DeFi. As the technology is gradually recognized and accepted, blockchain has entered the mainstream and is deeply integrated with the real world.

The negative side is that market competition has become increasingly pure and absurd. The lowering of the threshold for asset issuance means that the “dark forest” of the crypto market has become more dangerous. Today, it only takes a few clicks, along with some pictures and a few lines of text, to start a massive zero-sum game.

But why not channel this force in a positive direction and promote the development of the industry?

For example, some current AI Meme projects are beginning to gradually evolve into actual AI agents, rather than the past AI assistants that can only "talk nonsense". The recently popular DeSci (decentralized science) can also be regarded as a "scientific version of ICO". Although it is still meme-driven at the moment, in the long term, combined with the advantages of blockchain, DeSci can make traditional scientific research more transparent, easier to disseminate, easier to raise funds, and promote scientific research cooperation on a global scale.

However, it remains to be seen whether this model can be successful in practice, or how it will develop.

In fact, the idea behind DeSci is similar to what I mentioned in my GameFi-related article: How blockchain can make a difference in small studios short of funds and manpower. Under the circumstances, promote the development of independent games.

The core problems facing on-chain financing are:

•The threshold for asset issuance is too low< /p>

•Lack of restrictions

•Excessive financing capacity

This may be attributed to the extremely low entry barrier to blockchain. How to restrict the use of funds through rules and force project teams to continue to create truly valuable products is what we need to focus on.

Let players compete freely and builders continue to move forward. This is the prerequisite for the continued development of blockchain. In the next year, we may see more iterative versions of the “ICO” model, but what I am really looking forward to is the next “DeFi Summer” in this grand game.