Author: Zen, PANews

As one of the most anticipated AI native public chain projects this year, Sahara AI has attracted a lot of attention since its launch. From the enthusiastic investment of first-tier capital, BuildPad oversubscribed 8.7 times, to direct access to multiple top exchanges on the day of launch, to its performance on South Korea's Upbit, its trading volume once jumped to the second place on the platform, second only to BTC and XRP, which shows the market's high expectations and recognition of it.

Sahara is not only an important representative in the AI × Web3 track, but is also becoming a key coordinate in the narrative of the new generation of infrastructure. Therefore, in this interview, we specially invited Tyler, co-founder of Sahara AI, to provide in-depth answers to the most concerned issues of the outside world.

AI×Web3 needs to be built on real needs

PANews: Many people know that Sahara is a decentralized AI project. Backed by a top team, it has not only received a total of US$43 million in financing led by Polychain, Binance Labs, and Pantera Capital. The launch of Sahara token has also won support from global top exchanges such as Binance, OKX, and Upbit. What do you think is the core reason why Sahara has received so much recognition and support?

Tyler: I think the reason why Sahara has come this far is not because of popularity or capital, but because we have been standing on a real structural demand from the very beginning.

We have been asking ourselves: If AI is a deterministic future, what should its infrastructure look like? Who will build it? Who will participate? Today, most AI projects are still at the application layer, talking about how to bring AI to Web3, while Sahara wants to solve a more basic problem: Can Web3 become the entry point and benefit distribution mechanism for the AI era?

Sahara's answer is: Not only can we do it, but someone must do it.

The value distribution of AI cannot always be controlled by a few large model platforms. If AI in the future is really going to become an infrastructure like the Internet, then we need a more open, fair, and sustainable architecture. This is exactly what Sahara wants to do: solve a structural problem that we believe will be solved sooner or later.

We are not "doing some AI on Web3", but building a complete systemic operating platform for AI from scratch. We use the chain-level architecture to carry out the confirmation, call and profit sharing of AI, turning a system that used to rely on a large model platform for closed operation into a new network that everyone can participate in, build and share.

This is why we are the only AI Layer1 project that is jointly launched on OKX, Binance, Upbit, and Bithumb. It is not because we are at the forefront of the trend, but because we are really doing the difficult but right thing in the dimension of infrastructure.

PANews: There are many AI-related projects in the market. How is Sahara's positioning and path selection different from them? In your opinion, what is Sahara's biggest highlight or core advantage in the Web3 AI track?

Tyler: Many projects now use Web3 as a financing channel and AI as a publicity highlight, but this was not our starting point. What we really care about is: AI is becoming a new technological paradigm, so why should Web3 participate in it?

How does Web3 support the development of AI? I think that only with the support of AI can Web3 really have the opportunity to evolve into what it should be.

We have invested a lot of energy in studying how AI works: how models are trained, how data is called, and how value flows back to contributors. When we string these issues together, we will find a reality: today's Web3 still lacks an underlying system that can truly support AI collaboration.

So we simply started from scratch and designed a complete asset lifecycle system of "data → model → agent → income". Each link can confirm ownership, call, leave traces, and share profits, and all value flows can be clearly tracked and incentivized through on-chain logic.

This is like a "universal operating system" on which AI developers, contributors, and users can collaborate frequently. And this system is not just a concept on the architecture diagram. We have embedded it natively into our own L1 blockchain, serving AI scenarios with high frequency calls and complex interactions from the beginning.

For example, Ethereum allows us to freely build a financial system, from wallets, DEX to governance protocols, all can be completed on the chain. The same is true for Sahara, except that we are not serving financial assets, but moving the entire AI supply chain to the chain. Our niche is clear: to become the Web3 infrastructure of the AI era. This is Sahara's positioning, and we think we are the most core niche in this field.

With over 1.4 million daily active users, AI collaboration is not just empty talk

PANews: The SIWA test network has performed very well, with more than 3.2 million accounts and significant improvements in data labeling, task completion rate, accuracy, etc. What do you think of the value of these interim results to Sahara in building a decentralized AI infrastructure?

Tyler: SIWA is not just a "test network" for us, it is actually a systematic verification: Can the entire on-chain AI collaboration model we designed run? The result is that it can, and it exceeds our expectations.

Currently, there are more than 3.2 million wallets and 1.4 million daily active users on SIWA. What excites us most is that these users are the ones who actually do things on it - for example, more than 200,000 community data contributors have completed labeling, verification, and interaction on the chain, and there are millions of people waiting for the whitelist. This allows me to see the prototype of this AI resource collaboration network.

This incident also verifies a core judgment: AI collaboration is not just empty talk, but a real demand that already exists but has never been properly carried out. The greatest value of Web3 lies in the fact that it can provide a new order for this kind of multi-role, multi-round, high-frequency collaboration.

If SIWA is the first step for us to verify the on-chain collaboration capabilities, then the next step is the key node for us to truly "run through" the entire AI value cycle: we just launched the public beta of AI Agent Builder and AI Marketplace at the end of June. One allows everyone to create and publish AI Agents, and the other allows these Agents to be truly called, authorized, and profitable.

This can be seen as the prototype of the decentralized AI App Store and economic system. The infrastructure behind this is the entire chain-level architecture built by Sahara. In the future, we will also add capabilities such as profit sharing, on-chain task calling, and combined trading, so that the entire system can not only run, but also run for a long time.

So for us, these achievements are not just "stage successes", but also the first strong proof of the feasibility of decentralized AI. We hope that everyone can participate in AI, contribute to AI, and gain real value from it. And Sahara is bringing this wish into reality.

PANews: Compared with other public chains, Sahara has more products and a good product market fit. Many people say that Sahara's three-layer architecture is a new way to play "AI native chain". Can you talk to us about how this architecture helps the platform evolve step by step in terms of performance, governance, and ecology? If AI agents are put on the chain on a large scale or applications explode in the future, where will its expansion capabilities be reflected?

Tyler: AI itself is a system that relies heavily on collaboration and combinatorial development. It is impossible to run it just like deploying a smart contract. It involves a whole set of complex operations such as data uploading, model training, inference calling, and revenue tracking. Moreover, these operations also require rights confirmation, privacy protection, and multi-role collaboration, which places high demands on the performance, composability, and verifiability of the underlying chain.

So we developed a three-layer architecture: the bottom layer is the "infrastructure layer", the top layer is the "application layer", and the top layer is the "ecological coordination layer".

The infrastructure layer is our main chain, which cooperates with TEE (trusted execution environment) and on-chain contracts to register and confirm the ownership of data, models, and agents, and also ensure the privacy and security of the calling process. You can think of it as a "trusted registry + operating system" to solve the fundamental problems of who owns AI assets, how to use them, and how to share profits.

The application layer is the part where we are currently advancing the fastest. AI Agent Builder, Data Service Platform, and AI Marketplace are all here. Developers can upload data, build models, train, and even directly combine other people's components to create new functions. They can also trade, use, and monetize in the market. We are very "pragmatic" in this part, hoping that developers can really make a working Agent from 0 to 1 in a short period of time without having to mess around with various tools.

The ecological coordination layer is the most futuristic part and the source of our network effect. It is not only a technology, but also a connection mechanism. Web3 protocols, Web2 applications, computing service providers, and AI content creators can all access Sahara through standard interfaces. Only when everyone performs their duties and shares the benefits can a self-driven AI collaborative economy be created. Simply put, this layer is how we "pave the way" for the ecosystem.

Let's get back to the question, "What is Sahara's scalability?" We believe that in the future, whether it is the explosion of AI applications in Web2 or Web3, these assets will eventually need a chain to register, track and settle. The more applications and the more complex the combination, the more valuable Sahara will be. The purpose of designing this system has never been to support a single application, but to support the operation of an entire AI network. You can think of us as the "underlying operating system" of the AI world. Everyone can build, combine, and run models, and finally share the benefits. Once this kind of system is running, it will run faster and faster and become more powerful as it is used. Today we have just started, but you can already see some phenomena: more and more AI developers, model platforms, and even traditional enterprise data businesses are building their own on-chain AI products through Sahara, truly forming a compound interest effect.

What we see is not short-term user growth, but the entire AI industry's reliance on the chain in ten years. We want to lay a solid foundation for this.

The next step is to build a truly “usable AI collaboration network”

PANews: In the next 3 to 6 months, everyone is very concerned about what new products or collaborations Sahara will have, so that the outside world can see its key progress on the road of decentralized AI. Can you tell us in advance some highlights worth looking forward to?

Tyler: Our pace in the next few months is very clear: product, cooperation, mainnet, running this system step by step, and truly establishing a "usable AI collaborative network."

Let me introduce two core modules that we launched at the end of June: AI Agent Builder and AI Marketplace. Users can directly drag components, adjust models, and launch Agents without writing code, and the entire process is completed on the chain. You don’t need to jump to Github or Hugging Face to download things. We have modularized and integrated data sets and open source models. It can be said that now anyone who wants to make an AI application can open Sahara and do it directly.

This is a start. What we need to do in July is to make this system "run smoothly". The most important thing is to launch the monetization mechanism: model calling, API authorization, profit sharing, and License NFT will all be available this month. Allow users to use AI assets and earn profits through interaction. This is also the step that many early AI investors are most concerned about: how to turn models and data into assets and how to circulate in the ecosystem.

At the same time, we will also open the data service platform (DSP) from whitelist mode to full public beta, which everyone can participate in. This also means that from this stage on, everyone can build and benefit on Sahara.

Our mainnet will be launched in the fall, and the entire economic system will be officially closed. The chain will record all actions such as asset registration, model call, revenue distribution, user interaction, etc. During this process, $SAHARA will also be officially "activated" and become the core asset used when calling AI assets, paying fees, passively collecting authorized profits, and participating in platform governance.

In addition to products, we will also announce some ecological cooperation, including opening APIs to more AI tools, connecting the entrances and exits with tool platforms, realizing model circulation, and introducing more enterprise-level partners to participate in data and model collaboration. It can be imagined that Sahara is no longer just a technology platform, but a new AI collaborative ecosystem that is constantly emerging.

We are expanding Sahara from a "chain + platform" to a large system that can truly support the AI development ecosystem. In the future, you will see more roles find their place in this system: developers, data annotators, model authors, platform integrators, and AI users. We are building a collaborative structure and closed-loop benefits for these people.

PANews: Sahara has always emphasized data rights confirmation and data assetization. In the future, how will the value of on-chain data be truly released? Will it become a key to Sahara's commercialization? In addition, facing many similar AI data rights confirmation projects in the market, where do you think Sahara's core competitiveness lies?

Tyler: We believe that data is not just a consumable that is used once, but an asset that continuously generates value. Just like real estate can be rented out and works can be licensed, data can also be repeatedly called, combined, and used to train new models or agents.

The problem we want to solve is that data cannot just be "confirmed", but also "used and continuously monetized". A data set uploaded by a user is not thrown away after use, but can be continuously called, trained, synthesized for new purposes, and even used in combination by other developers. If all these behaviors are run on the chain, the use path of the data can be fully recorded, and the income distribution can be automatically carried out. This is not just about confirming ownership, but creating a truly fluid "data economy".

The data service platform (DSP) we designed is the entrance to this system. It not only allows you to participate in tasks and get rewards, but more importantly, it allows your data to truly become a traceable, reusable, and authorizable on-chain asset. And our protogenesis link + call layer logic also allows the use track of this data to be fully preserved.

So what is the difference between us and other data rights confirmation projects? I would say that they only deal with "ownership rights", while we deal with "usage rights" and "monetization paths". Their goal is to make it clear "who owns this", while our goal is "how to make this thing valuable and continue to be valuable".

This is also where we think the most potential for business can be unleashed. For corporate users, you don’t need to look for data from five platforms and seven sets of tools. We can help you connect, set prices, and obtain authorization in one stop. For individual users, data is no longer just “consumed” but can continue to make money for you. Once this system is up and running, it will be the most stable growth flywheel for our entire ecosystem.

So I often say that data is not Sahara's "entry point", but the foundation of our business model. It is the underlying logic that enables us to build an AI ecosystem for a long time, and it is also one of our core competitive advantages that sets us apart from other Web3 AI projects.

Sustainable incentives: Let “real use” drive the value cycle

PANews: The biggest question the market has about the incentive mechanism is the economic design of sustainable positive incentives after the project's TGE. In your opinion, how does Sahara design its economic system and incentive mechanism to keep it going for a long time and truly create a win-win situation for users and projects?

Tyler: This problem can be said to be a challenge that all Web3 projects cannot escape. In the early stage, users were attracted by subsidies, but once the popularity was over, everyone left. Sahara's approach is actually very simple: do not create artificial prosperity, and return to "real use" itself. The starting point of our design of the incentive mechanism is to think about a problem: if no one uses this system, it will naturally have no value; but as long as it continues to be used, someone must continue to benefit.

Therefore, we let the incentives come from usage and contributions. Our user roles are diverse: ordinary users, data contributors, agent builders, enterprises, model developers... Each type of person can find their own position in the ecosystem, not only can they use it, but also can benefit from it.

Like ordinary users, you can actually directly participate in data labeling tasks, experience AI agents, and create community content. This is not a way to subsidize traffic by cheating, but a behavior that truly brings data, user feedback, and content contributions to the platform. As long as your behavior is valuable, it will be recorded by the system and generate income.

As for developers, we provide a complete set of tool chains. Users can upload models, deploy, set up profit sharing, and then authorize others to use and receive a share in the Marketplace. Moreover, the content you upload can also be reused by others to generate more compound interest. Going further up, there are also corporate users and computing power providers, these more B-side roles can also find their own business space.

What is special is that one person can play multiple roles at the same time. You can be a data contributor, a deployer of AI assets, and even recommend other users to use your model. The more you do, the richer your roles and the more diverse your rewards. This "complex identity" income structure is a sustainable design that we particularly emphasize.

In the final analysis, what we want to do is: every real call, transaction, and collaboration can automatically trigger value returns. As long as the system is still in use, incentives can flow. This logic does not rely on subsidies to maintain popularity, but is driven by the positive cycle within the ecosystem. We believe this is the most feasible and healthiest direction for the combination of Web3 and AI.

PANews: In your opinion, what will be the relationship between AI Layer1 and traditional public chains in terms of ecological niches in the future? Is Sahara filling in the gaps, reconstructing, or opening up a new track?

Tyler: We believe that the relationship between AI Layer1 and traditional public chains is not one of growth and decline, but rather a co-evolution of different system tasks. Traditional public chains are good at handling general assets such as financial transactions, DeFi, and NFT, while AI Layer1 was born to take on another clear and foreseeable huge demand scenario: the confirmation, incentives, transactions, and collaboration of AI assets.

The asset form and operation mode of AI are completely different from traditional encrypted assets: it not only contains non-static assets such as data, models, agents, but also contains a large number of call records, behavior logs, and collaborative logic generated during operation. And this information must be carried by a native blockchain environment with a clear structure, reliable execution, and continuous incentives. We are not trying to "compete" with traditional public chains, but we are actively building a platform that allows data contributors, model developers, agent builders, computing service providers, and end users to collaborate organically when the existing system is unable to carry the needs of AI collaboration.

Of course, this system is also extremely complex, far more complex than existing public chains: it must ensure data privacy and verifiability; it must support complex interactive logic and ensure execution efficiency; and it must provide a clear profit-sharing and ownership confirmation path for each participant. These are all extremely challenging, but we think this is worth doing and someone must do it.

So in a sense, Sahara is neither a "filling in" nor a "reconstruction", but rather promoting a new system paradigm: AI-native blockchain. It is not an extension of a certain track, but rather opens up the infrastructure for a brand new collaborative network. We hope that this network is open, co-built, cross-chain, and that everyone can participate in and that it will continue to be a positive cycle.

PANews: As a leading project in Web3AI, how do you view the development direction of the entire AI track? What do you think are the biggest opportunities and challenges in the future?

Tyler: AI is rapidly evolving from a professional tool to a part of everyone's daily life. Just like the Internet and smartphones in the past, they have now become an indispensable infrastructure for everyone. We believe that in the future everyone will have their own AI model or agent to assist with daily tasks, and it will become an essential tool in our lives. This is the trend we are most optimistic about.

But to truly realize this vision, three key issues need to be addressed: first, how to enable AI to be controlled and deployed by individuals; second, how to ensure the privacy and transparency of data and interaction processes; and third, how to build a fair, open, and sustainable economic system.

These three points are exactly what Sahara is working on. We are not just making a toolbox, but building a future-oriented operating system that empowers all users, developers, researchers, and companies to create, use, and benefit from AI. We believe that the biggest opportunity of Web3 + AI is that it brings a new way of value creation: an open, collaborative, and everyone-can-participate AI network.

Sahara's role can be better understood using three metaphors:

We are the "AWS of decentralized AI", providing the underlying computing power, storage, calling and incentive systems, and providing infrastructure support for the entire Web3 AI world;
We are like the "Tesla of the AI world", redefining the production, operation and collaboration of AI assets with transparent, verifiable and tradable mechanisms;
We are also the "App Store for AI Agents", allowing developers to publish quickly and users to access them freely, building an ecosystem with powerful distribution capabilities and business models.

The biggest challenge is the complexity of this system: we must ensure the stability and security of the system while making it easy to use and participate in. It must be able to support tens of millions of users and applications to truly become the starting point for the integration of AI and Web3. We believe that this direction is worth long-term investment and construction. And Sahara is taking this first step.

Llama完成600万美元种子轮融资，Founders Fund和Electric Capital领投

Popular Articles