Looking towards 2026, the Archetype team is focused on identifying emerging trends, opportunities, a

Looking ahead to 2026, the crypto industry is at the beginning of a major transformation. As Bitcoin becomes increasingly common in global asset allocation with the involvement of institutional investors, stablecoin payments are being implemented in various countries through policies. AI is integrating deeply into the blockchain-based economy, and the narrative of the industry is being rewritten. The seven trends mentioned in Archetype’s latest forecast—from low-risk credit systems to modular infrastructure, to the rise of blockchain-based identities and creator economies—are taking shape rapidly.

Whether it’s the clarity of regulatory frameworks, renewed focus from investors, or efficiency gains brought about by new technologies, 2026 seems like a crucial year when crypto truly enters the mainstream economic system. What we need to consider next isn’t whether crypto will change the world, but at what speed and in what contexts it will be implemented.

Building applications on the blockchain finally makes sense.

By Aadharsh Pannirselvam

The principle is simple: blockchains designed, built, and optimized for specific use cases will perform exceptionally well. And next year, the best blockchains will be those built using basic components and principles.

The developers, users, institutions, and capital that are now joining the blockchain ecosystem differ from those in the past. They have their own cultural perspectives (i.e., definitions of user experience), and they place more emphasis on these cultures rather than abstract concepts like decentralization and resistance to censorship. In practice, this sometimes aligns with our current infrastructure, and other times it doesn’t.

For cryptocurrency applications that are designed for non-professional users, such as Blackbird or Farcaster*, the user experience is of particular importance. Some centralized design decisions that were considered heretical three years ago—such as the need for managed nodes, single-shard systems, and customized databases—now seem entirely reasonable. This is also true for stablecoin chains and trading platforms like Hyperliquid* and GTE, which operate on the basis of milliseconds, seconds, and optimal prices.

However, not all new applications are like this.

For example, the comfort derived from centralization is balanced by the growing concern among institutions and individuals regarding privacy issues. The needs and expectations of cryptographic applications can vary greatly, and their architectures should reflect this.

Fortunately, building a supply chain from scratch to meet these specific user experience requirements is no longer as complicated as it was two years ago. Nowadays, it’s almost as simple as assembling a custom computer.

Of course, you can choose any hard drives, fans, and cables you desire. But if you don’t need such a detailed configuration (which is quite likely), then you can opt for services like Digital Storm or Framework. These companies offer pre-configured computers that can be tailored to meet various needs. If you fall between these two options, you can still add your own components on top of those already available from these providers. This way, you gain greater modularity and flexibility. You can also remove any components you don’t need, while still ensuring that the final product maintains excellent performance.

By assembling and adjusting basic components such as consensus mechanisms, execution layers, data storage, and liquidity, applications can create forms that reflect unique cultural characteristics. These forms always reflect different user experiences, catering to distinct target audiences, thereby maintaining their value. The differences between these forms can be significant, such as between ToughBook, ThinkPad, desktop computers, and MacBooks. However, they also coexist to some extent—not every computer has its own independent operating system. More importantly, every necessary component becomes something that applications can freely iterate upon and adjust without worrying about breaking the overall protocol.

鉴于Circle已从Informal Systems收购了Malachite，显然，拥有定制区块空间的自主权已成为当前更广泛的优先事项。在未来一年，我期待看到应用程序和团队能够围绕Commonware和Delta等公司提供的基础组件和合理的默认值来定义和拥有其链资源，这类似于区块链和区块空间领域的HashiCorp或Stripe Atlas。

归根结底，这将使应用程序能够直接掌控自己的现金流，并利用其构建的表单的独特之处，按照自己的方式提供最佳的用户体验，从而构筑持久的护城河。

预测市场将继续创新(好吧，至少部分市场会如此)

by Tommy Hang

本轮周期中最受瞩目的应用之一便是预测市场。所有加密货币交易场所的周交易量均创下20亿美元的历史新高，这清楚地表明，该领域已朝着成为主流消费产品的方向迈出了重要一步。

这股势头带动了周边行业的众多项目，这些项目旨在与Polymarket和Kalshi等现有市场领导者形成互补或取而代之。但在各种炒作之中，如何区分真正的创新和噪音，最终将决定我们在2026年应该关注哪些方面。

From a market structure perspective, I am particularly interested in those solutions that can reduce price differences and increase the size of open positions. Although creating such markets still requires permission and proper screening, the liquidity of these markets remains insufficient for both market makers and traders. Through products like lending, we have the opportunity to improve the optimal routing system, various liquidity models, and the efficiency of collateral mechanisms.

The volume of transactions by category is also an important factor that determines which trading platforms are superior to others. For example, in November, over 90% of Kalshi’s transactions came from sports-related markets, highlighting the natural advantage some platforms have in obtaining favorable liquidity. In contrast, Polymarket’s transaction volume in cryptocurrency and political-related markets is 5-10 times that of Kalshi.

However, off-chain prediction markets still have a long way to go before they can be widely adopted. The Super Bowl in 2025 is a good example: just that event alone generated $23 billion in daily transactions on off-chain gambling markets, which is more than ten times the total daily volume of all on-chain markets combined.

To bridge this gap, it will require talented and innovative teams to address the core issues of prediction markets. I will keep an eye on these players over the next year.

The co-curators will expand the scale of decentralized finance

by Eskender Abebe

There are two extreme models in the asset management aspect of DeFi: pure algorithms (with hardcoded interest rate curves and fixed rebalancing rules), or pure human intervention (through risk committees and active fund managers). Agents, on the other hand, represent a third model: artificial intelligence agents (LLMs + tools + mechanisms) responsible for managing assets and implementing risk strategies in vaults, lending markets, and structured products. These agents not only execute predefined rules but also analyze risks, returns, and strategies.

Take the role of curators in the Morpho market as an example. They need to define collateral policies, loan-to-value ratios, and risk parameters in order to create income-generating products. Currently, there are limitations due to human involvement in this process. Agents can help streamline this process. In the near future, you’ll see agents competing directly with algorithmic models and human managers.

When will we see DeFi at step 37?

When I discuss artificial intelligence with crypto fund managers, I usually get two completely different responses: either that low-level logic models will automate all trading processes, or that they’re just theoretical toys that cannot survive in real markets. Both views ignore the fundamental changes that are happening in the architecture. Agents bring calm execution, systematic strategy implementation, and flexible reasoning capabilities into areas where human interference is likely to cause problems, while pure algorithms remain vulnerable. It’s more likely that agents will supervise and/or write the underlying algorithms, rather than replacing them. Low-level logic models act like architects, designing secure frameworks, while deterministic code remains slow and inefficient.

When the cost of advanced reasoning becomes extremely low, the most profitable “treasury” will not be the one that has the smartest people, but rather the one with the most powerful computing capabilities.

Short videos are becoming the default format for discovering (and ultimately purchasing) favorite content. TikTok Shop achieved over $20 billion in transaction volume in the first half of 2025, almost doubling compared to previous years. It’s gradually becoming the way people around the world consume entertainment content as a form of shopping experience.

To this end, Instagram has transformed Reels from a defensive feature into a profit-making tool. This format provides more exposure and is expected to account for an increasing portion of Meta’s advertising revenue in 2025. Whatnot has shown that real-time, personalized sales methods can achieve higher conversion rates than traditional e-commerce approaches.

The core concept is simple: people can make decisions more quickly when viewing content in real time. Every swipe counts as a decision-making point. The platform understands this well, so the boundaries between the information stream and the checkout process are disappearing. The information stream itself becomes a new point of sale, and every content creator becomes a distribution channel. Artificial intelligence further accelerates this transformation. It reduces video production costs, increases the amount of content available, and makes it easier for creators and brands to test ideas in real time. More content means greater potential for conversion, and platforms can improve purchase intentions by optimizing every second of the video.

Cryptocurrencies perfectly fit into this transformation. Faster content requires faster and more efficient payment methods. As shopping becomes frictionless and integrated directly into the content itself, we need a system that can handle small payments, distribute revenues in a programmatic manner, and track the complex chains of influence among all parties involved. Cryptocurrencies were precisely designed for such processes. It’s hard to imagine how the era of ultra-large-scale streaming commerce would have developed without them.

Blockchains will drive new developments in artificial intelligence.

By Danny Sursock.

Over the past few years, the focus of artificial intelligence has been on the billions of dollars worth of competitions between large corporations and startups. Decentralized innovators, however, have had to work behind the scenes, making incremental improvements.

But since people’s attention has shifted elsewhere, some crypto-native teams have made significant progress in areas like decentralized training and reasoning. This quietly growing revolution is now moving from prototype stages to actual testing and production environments.

Now, teams like Ritual*, Pluralis, Exo*, Odyn, Ambient, Bagel, etc., are ready to embrace the golden age of artificial intelligence. This new generation of competitors could have a transformative impact on the development of artificial intelligence.

By using models trained in globally distributed environments, we can break through scalability limitations and adopt new methods of asynchronous communication and parallel processing that have been proven effective at production scale.

新的共识机制和隐私原语的结合，使得可验证和保密推理成为链上构建工具包中非常现实的选择。

革命性的区块链架构将(真正意义上的)智能合约与富有表现力的计算结构相结合，从而简化使用加密货币作为交易媒介的自主人工智能代理。

基础工作已经完成。

现在的挑战是将这些基础设施基础扩展到生产层面，并证明区块链为何能够推动超越哲学、意识形态或拟物化筹款实验的根本性人工智能创新。

现实世界的资产将在现实世界中得到应用

by Dmitriy Berenzon

多年来，我们一直听到代币化的概念，但随着稳定币的主流化、流畅稳健的充值/赎回机制的出现，以及全球范围内监管透明度和支持的提升，我们终于看到了RWA(风险加权资产)的大规模应用。根据……RWA.xyz*截至撰写本文时，各类代币化资产的发行量已超过180亿美元，而一年前仅为37亿美元，我预计2026年这一增长速度将会加快。

值得注意的是，代币化和金库是RWA的不同设计模式：代币化创建链下资产的链上表示，而金库则在链上资本和链下收益之间建立桥梁。

I'm very pleased to see that tokenization and wallets can facilitate access to various physical and financial assets. This includes commodities like gold and rare metals, as well as private credits used for operating funds and payment financing. There's also public and private equity investments, along with various global currencies. We should also explore more unconventional areas. I hope to see eggs, GPUs, energy derivatives, wage withdrawals, Brazilian government bonds, Japanese yen, and so on being listed on the blockchain!

It’s important to note that this isn’t just about putting more things onto the blockchain. Instead, it’s about using public blockchains to transform how capital is allocated globally, making previously opaque, slow, and isolated markets more accessible, programmable, and liquid. Once these assets are on the blockchain, we can benefit from combining them with already existing DeFi technologies.

Finally, many of these assets will undoubtedly face challenges related to transferability, transparency, liquidity, risk management, and distribution. Therefore, infrastructure that can help address these challenges is equally important and exciting!

The rise of agent-driven products is just around the corner

by Ash Egan

The next generation of the internet will be less influenced by the platforms we browse, and more by the agents we interact with.

We all know that robots and agents are playing an increasingly important role in all forms of online activities. Roughly estimated, this proportion has now reached around 50%, including both on-chain and off-chain activities. In the world of cryptocurrencies, robots are increasingly involved in trading, management, assistance, contract scanning, and performing various tasks on our behalf—from token transactions to fund management, smart contract auditing, and game development.

This is an era of programmable, intelligent networks. Although we have been living in this era for some time, 2026 will be the year when cryptocurrency products begin to be designed with robots in mind, rather than humans (in a positive, free, and non-distopian manner).

The details are still being worked out, but I really hope that the time spent navigating between different websites can be reduced, allowing more focus to be placed on a simple chat interface through which we can manage on-chain robots. It could be thought of as a chat tool similar to Telegram, but with intelligent agents that are specifically designed for certain applications/tasks. These agents can develop and execute complex strategies, search for information and data that are relevant to me, and report transaction results, risks and opportunities that need attention, along with selected information. When I assign them tasks, they will track down opportunities, filter out all irrelevant information, and act at the right moment.

The chain already has the necessary infrastructure in place. With the default-enabled data graph and programmable micro-payments integrated with chain-based social graphs and cross-chain liquidity channels, we have everything we need to support a dynamic proxy ecosystem. The plug-and-play nature of cryptocurrencies means that proxies can simplify complex processes and eliminate obstacles. The advantages of blockchain in this regard far exceed those of Web 2 infrastructure. This point cannot be overstated.

This may be the most important aspect. It’s not just about automation; it’s about being freed from the isolation of Web 2.0, eliminating friction and waiting times. We can already see this transformation happening in the search industry: approximately 20% of Google search results now include AI-generated summaries. Data shows that when users see these summaries, their likelihood of clicking on traditional search result links decreases significantly. Manual browsing of web pages becomes increasingly unnecessary. A programmable intelligent network will further expand this trend to the applications we use every day. I believe this is a positive development.

In this era, we will reduce our exposure to negative news, minimize panic-driven transactions, and time zone differences will diminish (there’s no need to “wait for Asia to wake up”). For every developer and user, interaction with the chain-based world will become more convenient and expressive.

As more assets, systems, and users join the chain, this cycle will continue to intensify.

The more opportunities available on the chain, the more proxies can be deployed, and the more value can be created. And so on and so forth.

但是，我们现在构建什么，以及如何构建它，将决定这个智能网络是会变成一层薄薄的噪音和自动化，还是会引发赋能型和动态产品的复兴。