The Integration of AI and Web3: Building Infrastructure for a New Era

In an era where technological paradigms are truly shifting, we often see the hype first, rather than the system. The current wave of AI is no exception. As a primary investor, I have always believed that betting on the deep transformative forces within the industry is far more valuable than chasing superficial narratives.

In the past year, I have come into contact with a large number of projects related to RWA, consumer applications, information finance, etc., all of which are exploring the intersection of the real world and on-chain systems. However, an increasingly obvious trend is that regardless of which route the project chooses, it ultimately needs to incorporate AI collaborative logic to enhance competitiveness and efficiency.

For example, RWA projects need to consider how to leverage AI for risk control optimization, off-chain data verification, and dynamic pricing; consumer applications and DeFi projects require AI for user behavior prediction, strategy generation, and incentive distribution. Whether it's asset digitization or experience optimization, these seemingly independent narratives will ultimately converge on the same technological logic: if the infrastructure lacks the integration and carrying capacity of AI, it will not be able to support the complex collaboration of the next generation of applications.

The future of AI is not just about becoming more powerful and widely applied; the real paradigm shift lies in the reconstruction of collaborative logic. Just like the early transformations of the internet, AI is moving towards a new stage: intelligent agents will become co-creation partners for everyone, helping to transform expertise, creativity, and tasks into automated productivity tools, and even enable value realization.

This is a problem that is difficult to solve in the current Web2 world, and it is also the core logic behind my focus on the AI+Web3 track: making AI collaborative, transferable, and profit-sharing is the system that is truly worth building. In this context, I want to discuss a project that, so far, is the only one attempting to systematically construct the underlying operations of AI from a chain-level structure: Sahara.

The Essence of Investment: Choosing a Value System

My investment logic is not simply to combine the public chain narrative with AI and then choose well-backed teams to bet on. Investment is essentially a choice of worldview, and I have been questioning a core issue: Can the future of AI be co-owned by more people? Can it leverage blockchain to reconstruct the value attribution and distribution logic of AI, allowing different roles such as ordinary users and developers to have the opportunity to participate, contribute, and continuously benefit?

In order to find the answer, I extensively researched various AI projects until I encountered Sahara. The response from Tyler, the co-founder of the project, was: to build an open, participatory ecosystem that everyone can own and benefit from. This simple answer precisely hits the soft spot of traditional public chains: they often serve developers in a one-sided manner, and the token economic design is mostly limited to fuel fees or governance, making it difficult to truly support a positive cycle of the ecosystem, and even harder to sustain new emerging tracks.

This road is full of challenges, but precisely because of that, it represents an irresistible revolution - which is also the reason I am firmly investing. The real paradigm shift is not about creating a single product, but about building a supportive system. Sahara is one of the most anticipated cases I expect.

From Investment to Additional Investment: Valuing Team Strength

I initially invested in Sahara because it is committed to building AI economies and infrastructure systems, which I believe is the true mission of AI. What led me to make an additional investment at an 8x pre-round valuation in just six months is the rare strength I felt in this team.

The two co-founders each have their own unique characteristics. One is the youngest tenured professor at the University of Southern California, specializing in the field of AI. In the year and more that I have known him, I have witnessed his qualities of working more than ten hours a day, maintaining emotional stability, and being humble. The other, Tyler, was the North American Investment and Incubator Director at a certain investment platform and has a deep understanding of Web3. His self-discipline is astounding: he strictly controls his sleep schedule, persists in fitness, avoids sugar to keep his mind clear, and works over 13 hours a day.

When someone says that Sahara has gained capital favor due to luck, I would frankly say that this is an inevitable result. In the current market where primary financing is difficult, Sahara is still sought after by investors. In addition to well-known investment institutions, some funds and banks that focus on AI have also shown strong interest in Sahara. This reflects a positive evaluation of Sahara's technical depth, team background, system design, and execution capability.

Sahara has already demonstrated some substantial achievements: the test network has activated over 3.2 million accounts, there are more than 200,000 data platform annotators (with millions more in queue), clients include several leading companies, and it has achieved revenue in the tens of millions of dollars. In terms of building the infrastructure chain, Sahara has progressed deeper and more steadily than most "AI concept projects."

The Ultimate Challenge of Public Blockchains: Achieving Continuous Benefits for Contributors and Positive Economic Cycles

In systems that combine AI and blockchain, a key issue is whether there is a mechanism that can recognize, record, and continuously reward every contributor. Model training and data optimization rely on a large amount of labeling and interactive support; conversely, a lack of user contributions will increase project costs and weaken the value-driven aspect of community co-construction.

Sahara is one of the few Web3 AI projects that allows ordinary users to participate in data construction from the very beginning. Its data labeling task system operates daily, with a large number of community users actively participating in labeling and prompt creation, which not only helps improve the system but also invests in the future.

Through the mechanism of Sahara, not only has the quality of the model improved, but it has also allowed more people to understand and participate in this decentralized AI ecosystem, linking data contributions with rewards to create a virtuous cycle. For example, a certain on-chain project quickly built a high-quality dataset using Sahara's decentralized data collection and human-machine collaborative labeling, significantly enhancing its model training efficiency, while users participating in data labeling also received token rewards.

Sahara's "permissionless copyright" mechanism ensures the protection of all participants' rights while guaranteeing the open circulation and reuse of AI assets, which is the underlying logic driving the explosive growth of the entire ecosystem.

Scenarios supported by long-term value

Imagine if you're building an AI application, you naturally want your model to be more accurate and closer to real users than others. The key advantage of Sahara is that it connects to a vast and active data network, including hundreds of thousands, and potentially millions in the future, of annotators. They can continuously provide you with customized, high-quality data services, allowing your model to iterate faster.

More importantly, this is not a one-time transaction. Through Sahara, you are connecting to a potential community of early users; and these contributors are likely to become real users of your product in the future.

Through the smart contract system and rights confirmation mechanism of Sahara, a long-term, traceable, and sustainable incentive system has been achieved. Regardless of how many times the data is called, contributors will receive continuous profit sharing, with earnings dynamically linked to usage behavior.

Sahara builds an economic system that covers the entire lifecycle of AI models, with a built-in profit-sharing mechanism at every stage, including model deployment, invocation, combination, and cross-chain reuse, allowing value to be captured over a longer period. Model developers, optimizers, validators, and computing power contribution nodes can continuously benefit at different stages, rather than relying solely on one-time transactions or buyouts.

This system brings a compound effect for model composition calls and cross-chain reuse. A trained model can be repeatedly called and combined by different applications, with each call creating new revenue for the original contributor.

Therefore, I agree with the underlying philosophy of Sahara: a truly healthy AI economic system should not just be about the plundering of data and the buyout of models, nor should it allow only a few people to reap all the benefits. It must be open, collaborative, and mutually beneficial—everyone should be able to participate, and every valuable contribution should be recorded and continuously rewarded in the future.

Challenges Faced

Although I am optimistic about Sahara, I will not overlook the challenges that the project will face due to my investment position.

One of the major advantages of the Sahara architecture is that it is not limited to any specific chain or single ecosystem. Its system was designed from the beginning to be open, cross-chain, and standardized: supporting deployment on any EVM-compatible chain while providing standard API interfaces that allow various Web2 systems to directly call Sahara's model services and complete on-chain settlements.

However, although this architectural design is extremely rare, it also carries core risks: the value of the infrastructure lies not in "what it can do," but in "who is willing to do what based on it." To become a trusted, adopted, and composable AI protocol layer, the key for Sahara lies in how ecosystem participants assess its technological maturity, stability, and future predictability. Although the system itself has been built, whether it can truly attract a large number of projects to land based on its standards remains unknown.

Sahara has achieved key validation: providing relevant data services for several leading enterprises and addressing some of the industry's most challenging data demand issues, which serves as an early signal of the system's feasibility. However, these collaborations mainly originate from the Web2 world. The long-term development of Sahara will ultimately depend on the maturity and penetration of the entire Web3 AI space. While Sahara benefits from the overarching trend of Web3 AI, it still requires the implementation and refinement of more Web3 native AI products and technological solutions to truly unlock the value of its infrastructure.

It is worth noting that Sahara is currently "one of a kind". In the arena of chain-level infrastructure designed for AI, although there are imitators proposing conceptual frameworks, only Sahara has fully realized the entire implementation from on-chain rights confirmation, off-chain execution, cross-chain invocation to a technical closed loop and real income, and has gained validation from actual customers.

This gives Sahara a "competitive advantage," but it also brings structural risks: once successful, it will define the industry benchmark for the entire Web3 × AI infrastructure; however, if it fails, AI Layer1 may be seen as a premature layout. As the only option in this field currently, the market's judgment on it is naturally harsher and more rational - it must withstand the test of time and ecology.

Advice for all builders and observers

For me, the core of every level one investment decision rests on three points: the depth of understanding of the world, the dimensionality of trend judgment, and the willpower of the team to navigate through cycles. Products and features are certainly important, but they often merely represent a concrete manifestation of these underlying cognitions.

Web3 does not lack ideas, nor does it lack stories; what it lacks are practitioners who can turn logic into order, and those who truly know what should be upheld and what should be abandoned.

I cannot guarantee that Sahara can become the next paradigm-level chain. But it is indeed the only attempt worth serious consideration, observation, and investment at present.

If you wait for the day when everything is ready, the ecosystem is formed, and industry consensus is established - then the opportunity may have already passed.

Perhaps you should feel a sense of urgency. Not because you missed something, but because you happen to encounter a time point when a system has just started.

While others are still observing and waiting for clear signals from the market, you already know that this system exists, the direction is clear, and the structure has been built, it just has not yet been widely understood.

Most people will flock to it after it succeeds, while you are currently standing at this critical juncture where the flywheel has not yet started and the standards have not yet been established.

This is not a deterministic opportunity, but it is a real beginning. Not everyone can understand it.