AI × Web3: Who will build the chain for this era?

When a technological paradigm truly shifts, we often first see the hype rather than the system. The AI wave we are experiencing is no different.

As a primary investor, I have always believed that betting on the transformative forces deep within an industry is far more worthwhile than chasing superficial narratives.
In the past year, I have seen numerous projects in RWA, Consumer, infoFi, etc. — all of which are undoubtedly exploring the intersection of the real world and on-chain systems.
However, an increasingly obvious trend is that regardless of the route a project takes, it ultimately must enter the collaborative logic of AI, leveraging AI to enhance competitiveness and efficiency.
For example, in RWA, thinking about how to use AI for risk control optimization, off-chain data verification, and dynamic pricing is the future; or in Consumer or DeFi, which urgently need excellent user experiences, AI is also required to accomplish user behavior prediction, strategy generation, incentive distribution, and other similar directions in various tracks I won't elaborate on further.
Thus, whether it is asset digitization or experience optimization, these seemingly independent narratives will ultimately converge into the same technological logic: if the infrastructure does not possess the ability to integrate and support AI, it cannot sustain the complex collaboration of the next generation of applications.
In my view, the future of AI is not just about being "stronger and more widely used," but the real paradigm shift lies in the reconstruction of collaborative logic. Just like the early transformations of the internet, it was not because we invented DNS or browsers, but because it allowed everyone to participate in content creation and turn ideas into products, thus giving rise to an entire open ecosystem.
AI is also on this path: Agents will become everyone's intelligent co-creator, helping you turn expertise, creativity, and tasks into automated productivity tools, even monetizing them.
This is a question that the current Web2 world finds hard to answer, and it is also some of the underlying logic I see in the AI + Web3 track: making AI collaborative, circulatable, and profit-sharing is the truly worthwhile system to build.
What I want to talk about today is the only project so far that attempts to systematically construct the underlying operations of AI from a chain-level structure: Sahara.
(Note: I am not from the Sahara team, but as an investor in Sahara for two consecutive rounds, I have accompanied the project for nearly two years, witnessing advantages and potential beyond the public perspective, but I inevitably carry subjective expectations.)

The essence of investment is a worldview, a value system of choice

My investment logic is not just about the narrative of public chains combined with AI, then seeing which team seems to have a better background before placing a bet.
Investment, at its core, is a choice of worldview, and I have always been asking a core question: Can the future of AI be jointly owned by more people?
Can it leverage blockchain to reconstruct the value attribution and distribution logic of AI, allowing ordinary users, developers, and other roles to have the opportunity to participate, contribute, and continuously benefit? It's simple: only with the emergence of this logic do I believe such projects have the potential to be disruptors, rather than just "abandoned public chain +1." To find the answer, I have basically scanned all the AI projects I could access until I encountered Sahara. The answer given to me by Sahara's co-founder Tyler was: to build an open, participatory ecosystem that everyone can own and benefit from.
This statement is simple, but it precisely hits the soft spot of traditional public chains: they often serve developers unidirectionally, and token economic designs are mostly limited to Gas Fees or governance, rarely able to truly support the positive cycle of the ecosystem, let alone sustain the development of an emerging track.
I am well aware that this path is full of challenges, but precisely because of this, it is a revolution that cannot be refused — and this is why I am firmly investing. Just as I emphasized in my previous article discussing the "evolution from Web2 to Web3": the real paradigm shift is not about creating a single product, but about building a supportive system. (Readers interested in this logic are welcome to refer to that article.)
And Sahara is one of the most anticipated cases in my previous predictions.

From investment to an 8x valuation follow-up

If I initially invested in Sahara because it is doing what I believe to be the true leading mission of AI — building AI economic and infrastructure systems, then the reason I rushed to follow up with an 8x investment in just six months is that I felt an exceptionally rare strength in this team.

• Two co-founders: one is the youngest tenured professor at USC, specializing in AI. The value of a tenured professor in a U.S. university in their 90s, in my opinion, is not just in academia, but in the fact that at this age, they still have dreams, energy, and the determination to realize those dreams. Knowing Professor Ren for over a year has shown me what it means to work for more than ten hours a day, with stable emotions and humility.
• Tyler, former investment director at BN Lab, responsible for North American investments and incubators, his understanding of web3 goes without saying. He is astonishingly disciplined: he only sleeps in multiples of 1.5 hours, insists on exercising to maintain his condition no matter how busy he is, and avoids sugar to keep his mind clear, working over 13 hours a day. I once joked that he is a robot, and he simply replied, "I am lucky to have this busy life today." His source of dopamine is pushing the progress of the project every day; dream-building is his passion, needing no other fuel.

I am grateful to have met them; it has changed me. I have also finally started to maintain a regular sleep schedule, stabilize my emotions, and exercise…
So when someone says Sahara gained the favor of capital due to luck, I will unreservedly add, "the pursuit of capital is a natural result." I vividly remember how difficult it was to raise capital in this round, but Sahara was being chased by the primary market for investment.
What everyone remembers is that Polychain, Binance, and Pentera invested in Sahara. Sahara has opened the investment era for Samsung's entry into the Web3 AI field, and its receipt of the Samsung AI Award was a significant reason for Samsung's investment. In addition, several AI-heavy funds, national banks, and others are also guests of Sahara. You can see a group of institutions more inclined towards traditional technology and industrial resources quietly betting on AI × Web3 because of Sahara.
Capital will only pay for certainty in direction and execution — this is a positive feedback on Sahara's technical depth, team background, system design, and execution capability.
This is also why it can achieve some real and solid structural indicators: over 3.2 million accounts have been activated on the test network, with over 200,000 data platform annotators (millions are in line), and their clients include leading companies like Microsoft, Amazon, Character.AI, Motherson, etc., already achieving millions of dollars in revenue.
On this infrastructure chain, at least from "who will do it" to "can it be done," Sahara has already gone deeper and steadier than 99% of "AI Narrative projects."

The ultimate question for public chains: enabling all contributors to continuously benefit and drive positive economic cycles

Returning to our initial judgment logic: in the system combining AI and blockchain, is there really a mechanism that allows every contributor to be seen, recorded, and continuously rewarded?
Model training and data optimization rely on a large amount of annotation and interaction support; conversely, if there is a lack of user contributions, the project itself must invest more funds to procure data and outsource annotations, which not only increases costs but also weakens the value-driven community co-construction.
Sahara is one of the few Web3 AI projects that allows ordinary users to "participate in data construction from day one." Its data annotation task system operates daily, with a large number of community users actively participating in annotations and prompt creation. This not only helps improve the system but also invests in the future with data.
Through Sahara's mechanism, it not only enhances model quality but also allows more people to understand and participate in this decentralized AI ecosystem, linking data contributions with rewards, forming a true virtuous cycle.

• A typical example is the Myshell project on the BNB Chain, which quickly built a high-quality dataset covering multiple languages and accents by leveraging Sahara's decentralized data collection and human-machine collaborative annotation, significantly improving the training efficiency of its TTS and voice cloning models. This also propelled its open-source projects VoiceClone and MeloTTS to gain thousands of GitHub stars and over 2 million downloads on Hugging Face.

At the same time, users participating in data annotation also received token rewards issued by Myshell, forming a two-way incentive loop between developers and data contributors.
Sahara's "permissionless copyright" mechanism ensures the rights of all participants while guaranteeing the open circulation and reuse of AI assets — this is the underlying logic driving the explosive growth of the entire ecosystem.
Why is this a scenario with long-term value support?
Imagine if you want to build an AI application, you naturally hope your model is more accurate and closer to real users than others.
Sahara's key advantage is that it connects you with a vast and active data network — hundreds of thousands, potentially millions 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 early user community; and these contributors are likely to become your product's real users in the future.
This connection is also not a one-time buyout; through Sahara's smart contract system and rights confirmation mechanism, it enables a long-cycle, traceable, and sustainable incentive system.

• No matter how many times the data is called, contributors will receive continuous profit-sharing, with earnings dynamically linked to usage behavior.

But this is not just a revenue model for data annotation and model training stages. Sahara is building an economic system that covers the entire lifecycle of AI models, with built-in profit-sharing mechanisms at every stage of model deployment, combination, and cross-chain reuse, allowing value to be captured over a longer period.
Model developers, optimizers, validators, and computing power contribution nodes can all continuously benefit at different stages, rather than relying solely on one-time transactions or buyouts.
Such a system brings a compounding effect to model combination calls and cross-chain reuse. A well-trained model can be repeatedly called and combined by different applications, creating new earnings for the original contributors with each call.
For this reason, I resonate with Sahara's underlying belief: a truly healthy AI economic system cannot just be about data plundering and model buyouts; it cannot just allow a few people to reap all the benefits. It must be open, collaborative, and win-win — where everyone can participate, every valuable contribution can be recorded, and continuous rewards can be obtained in the future.

But the closer we get to the real structure, the more challenges arise

Although I am optimistic about Sahara, I will not hide the challenges the project will face due to my investment stance.
One major advantage of Sahara's architecture is that it is not limited to a 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 also providing standard API interfaces, allowing Web2 systems — whether e-commerce backends, enterprise SaaS, or mobile apps — to directly call Sahara's model services and complete on-chain settlements.
However, despite this architecture being extremely rare, it also carries a core risk: 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, Sahara's key lies in how ecological participants assess its technical 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 uncertain.
Undeniably, Sahara has achieved key validations: serving leading companies like Microsoft, Snapchat, MIT, Motherson Group, Amazon, etc., providing them with relevant data services and addressing some of the industry's most challenging data needs, becoming an early signal validating the feasibility of this system.
However, it is important to note that these collaborations mainly come from the Web2 world; what truly determines Sahara's long-term development is still the maturity and penetration of the entire Web3 AI track. Sahara benefits from the Web3 AI big trend, but to truly release the value of its infrastructure, it still needs more Web3 native AI products and technical solutions to land and improve.
But don't forget, Sahara is currently "the only one of its kind."
In the chain-level infrastructure track designed specifically for AI, while there are no shortage of imitators proposing conceptual frameworks, only Sahara has achieved comprehensive landing from on-chain rights confirmation, off-chain execution, cross-chain calling to technical closure and real revenue, validated by actual clients.
This not only gives Sahara a "monopoly advantage," but also brings structural risks: once successful, it will define the industrial benchmark for the entire Web3 × AI Infra; but if it fails, it may be seen as a premature layout for AI Layer1.
Since it is currently the only choice in this field, the market's judgment on it will naturally be more stringent and rational — it must withstand the test of time and ecology.

Finally, a message to all builders and observers | Seize the window of the building period, rather than regret it after it has taken shape

For me, the core of every primary investment decision rests on three things: the depth of understanding of the world, the dimensions of trend judgment, and the willpower of the team to traverse cycles. Products and functions are certainly important, but they often only reflect these underlying cognitions concretely.
Web3 is not short of ideas, nor stories; what it lacks is the hands to turn logic into order, and what it truly needs are people who know what to persist in and what to abandon.
I cannot guarantee whether Sahara can become the next paradigm-level chain. But it is indeed the only attempt worth taking seriously, observing closely, and investing in at present.
If you are waiting for the day when everything runs smoothly, the ecosystem takes shape, and industry consensus is established — then the opportunity has already passed you by.
So, perhaps you should really panic. Not because you missed something, but because you just happened to hit a moment when the system is just beginning. While others are still watching and waiting for the market to give clear signals, you already know that this system exists, the direction is clear, and the structure is in place, only that no one truly understands it yet. Most people will swarm in after it runs smoothly, while you — at this moment — are standing at a node where this flywheel has yet to start, and the standards have yet to be defined.
This is not a certain opportunity, but it is a real beginning. Not everyone can understand it, but you have already seen "something ahead of consensus."