X SPACES: ARKREEN/TALK X STRATOS AMA🔗︎
Introduction🔗︎
Alright, yes, we’re all set to begin. Honestly, I find these online discussions incredibly exciting—sometimes even more engaging than in-person events. It feels like attending a masterclass every week, where I get to hear from some of the leading experts in DePIN and AI from around the world. It’s truly inspiring, and I’m sure many in our community feel the same way.
Today, I’m especially thrilled to welcome a very special guest—Mr. Bin, the founder of Stratos Network. Let’s give him a warm welcome! Mr. Bin, it’s a pleasure to have you here with us on the Live Space. Thank you for joining us! How’s the weather in your city today?
Wow, it’s incredible! I’m currently in Montreal, Canada, and you wouldn’t believe it—last week, we had nearly a meter of snowfall! It was the heaviest snowfall in over 100 years, breaking a century-old record. It’s been absolutely crazy. In fact, right now, I’m sitting in my car, stuck on the street because of the snow. But no worries—we can still have a great conversation!
tay safe out there! That sounds incredible—I can only imagine how beautiful the city must look covered in snow. Here in Singapore, our office is always chilly, so I’m bundled up in a thick jacket just to keep warm.
Alright, let’s dive in. To start things off, could you share a bit more about yourself? Specifically, what inspired you to create the Stratos Network, and how do you see it fitting into the broader Web3 landscape? We’d love to hear your thoughts. Thank you!
My name is Bin Zhu, and I’m the founder of Stratos. I’ve been working in the tech industry since 2000—more than 20 years now. Over the years, I’ve worked with several major companies, and during my time at one of the largest traffic-heavy websites in the world, I had a realization.
I started thinking about the key differences between Web2 and Web3. In Web2, we have dominant companies like Facebook, Google, and LinkedIn—huge social networks built on solid infrastructure. These platforms rely on well-established cloud services such as Google Cloud, AWS, and Azure. In China, companies like Alibaba and Tencent provide similar foundational services.
However, in Web3, we don’t yet have that kind of robust infrastructure. We have strong foundational technologies like Bitcoin and Ethereum, but they primarily serve as ledgers. Even newer public chains like Solana focus on ledger functionalities rather than broader infrastructure needs. The Web3 ecosystem is far more complex, with data existing in multiple formats, requiring a strong foundation to support its growth and scalability.
That’s what inspired us to build Stratos. We saw an opportunity to leverage our knowledge from Web2 and apply it to Web3, creating the foundational infrastructure necessary for a truly scalable decentralized internet.
In 2021, we officially launched Stratos, issued our tokens, and began building the Stratos Network. Since then, our focus has been on laying a strong foundation.
When constructing a building, the foundation is always underground—out of sight—while the visible structure above ground is what people notice. The same applies to Stratos. Over the past few years, we’ve been working behind the scenes to establish the core infrastructure necessary for a scalable and efficient network.
Today, we’ve successfully built our blockchain and storage network because we want people to truly experience what Stratos has to offer. Just last week, we launched Stratos Wallet, which functions similarly to a decentralized Dropbox or Google Drive. Anyone with a Stratos Wallet can try out our decentralized storage solution. As long as users hold STOS tokens to cover traffic fees, they can access and interact with the network seamlessly.
I’d be happy to dive into more details as our conversation continues. Thank you!
Stratos PoT🔗︎
That’s fascinating. One of the standout features of Stratos is its unique Proof-of-Traffic (PoT) consensus mechanism. Could you explain how it works and how it sets Stratos apart from other blockchain projects focused on storage and data management?
Over to you, Mr. Bin.
Stratos Storage is an idea we developed after studying existing storage-based blockchain projects. For example, take Filecoin—many projects use a Proof-of-Storage mechanism, which essentially means that the more storage you provide, the more rewards you earn. While this makes sense in theory, it presents a major issue: storage alone doesn’t necessarily create value for the community or users.
Here’s the problem: If I’m a miner and my only concern is storing data to earn rewards, I could simply generate meaningless files—empty 2GB or 1TB files—and store them on my local disk. The system would recognize that I’m storing large amounts of data and reward me, even though that data is completely useless. This is a fundamental challenge for many decentralized storage projects—it’s difficult to verify whether the stored data is actually valuable or not. Miners, in many cases, don’t care if the data is fake or garbage, as long as they continue earning rewards.
That’s why we believe Proof-of-Storage is flawed and decided to take a different approach: Proof-of-Traffic (PoT). Instead of rewarding miners just for storing data, we focus on the actual usage and transfer of that data. Simply having a large disk space isn’t enough—you also need to provide meaningful content that users want to access.
Think about platforms like YouTube, Netflix, or TikTok. They don’t just store videos—they enable people to stream and interact with content that interests them. Similarly, in Stratos, miners can’t simply fill their storage with random files and expect to earn rewards. If no one accesses their data, they won’t generate traffic, and without traffic, they won’t earn rewards. This ensures that the storage network only holds meaningful and valuable data.
Additionally, PoT introduces another important factor: network performance. Miners can’t just focus on having large disks; they also need to ensure they have good bandwidth and fast server performance. If their connection is too slow, even if they store high-quality data, users won’t be able to access it efficiently. So, miners must optimize both the quantity and quality of their storage, as well as the speed and reliability of their service.
This creates a self-sustaining and efficient ecosystem where miners are incentivized to provide real value. Similar to how Bitcoin’s security model makes attacking the network unprofitable, Stratos ensures that bad actors can’t game the system. Instead, our PoT mechanism encourages positive behavior, leading to a stronger and more effective decentralized storage network.
That’s the power and beauty of Proof-of-Traffic in Stratos.
Stratos Milestones🔗︎
Thank you! That’s truly innovative.
Let’s talk about Stratos’ journey so far. What are some of the major milestones the project has achieved since its inception? And looking ahead, what key milestones should we be watching for in 2025?
We’d love to hear your insights!
Stratos has come a long way, and we’ve achieved several major milestones over the years.
-
First, we successfully launched the Stratos blockchain, which has been running smoothly for almost two years. This forms the core of our decentralized ecosystem.
-
Second, we deployed our storage network, which has now been operational for over a year. It consists of around 1,000 nodes distributed across more than 20 countries, providing a truly decentralized and resilient infrastructure.
-
The third major milestone was the recent launch of Stratos Drive just last week. Now, users don’t need any technical knowledge or run their own nodes to use our decentralized storage. With just a Stratos Wallet, they can upload, download, and share their data securely—whether for personal use, among friends and family, or in business environments.
Another important milestone we just hit is the all-time high number of active mining nodes on our network, now surpassing 1,100. This is a sign of growing participation and decentralization, and we expect even more nodes to join in the near future.
Looking ahead, the next major milestone will be
-
The integration of IPFS with Stratos, which we plan to launch in just two weeks. While many projects and even Web2 companies use IPFS, it’s important to note that IPFS itself is just a protocol—it doesn’t inherently provide a fully decentralized storage solution. However, with Stratos, IPFS will run on a completely decentralized network, making it the largest IPFS-powered storage network in the world, with over 1,000 participating nodes and unmatched storage capacity and efficiency.
-
Later this year, we’ll also introduce a decentralized video platform—essentially a decentralized YouTube. Users will be able to upload and share videos, and unlike YouTube or TikTok, where platforms control content, Stratos ensures that creators retain full ownership. If content is stored on our network, it can’t be taken down due to regulations or platform policies. Additionally, through our Proof-of-Traffic mechanism, content creators will be rewarded with Stratos tokens based on the amount of traffic their videos generate. The more views they get, the more they earn.
Beyond that, we have two more ambitious milestones:
-
Decentralized Database for AI – AI applications require databases, but not traditional relational databases. Instead, they rely on vector databases, which we are actively developing. This year, we will release a test network for a decentralized vector database on Stratos. It will allow users to store both structured and personal data with extremely fast response times—measured in milliseconds rather than seconds or minutes.
-
Deploying DeepSeek AI on Stratos – DeepSeek is a powerful open-source AI model, and our team is currently working on deploying it on the Stratos network. The key difference from traditional AI services is privacy. Most AI models collect user data during interactions, but with Stratos, user data remains fully private and decentralized. Users will own their own AI-generated data, ensuring it isn’t shared or monetized without their consent.
2025 will be an incredibly exciting year for Stratos. We’re moving beyond just laying the foundation—we’re actively building and delivering real, usable products that our community can see, try, and experience firsthand. Our focus is on making Stratos not just a strong infrastructure but a platform that people can truly benefit from in their everyday digital interactions.
Stratos Scalability🔗︎
Yes, that sounds promising. Traders are aiming to establish a decentralized AI infrastructure, and topics such as scalability, security, and decentralization are always key considerations.
How does Stratos address the challenges that other networks typically face in these areas? Thank you.
Stratos operates with three distinct models.
The first is the Stratos blockchain.
The second is the Stratos Meta Node layer, which functions as the brain of the system. This layer optimizes and accelerates traffic, allowing for efficient data transfer across various locations—whether from the US to China or from Hong Kong to Europe. This component is crucial because it enables us to deliver high-performance storage services, which is also why we partner with several AI companies. I can share more details on that later.
The third model is the storage layer, where real user data is stored on resource nodes. This layer is fully scalable, and while we currently have over a thousand nodes participating, we foresee no issues with scaling to 10,000 or even 100,000 nodes. Ultimately, we envision these resource nodes being distributed widely, allowing participation from various communities and neighborhoods.
When fully realized, Stratos will establish a purely decentralized network, functioning also as an edge network. This means that when you need to retrieve data, you may find that a Stratos node is already present in your neighborhood, eliminating the need to access a distant node. Importantly, the owner of the node won’t have access to the data stored there, ensuring privacy. In essence, these three layers are each dedicated to their specific functions, yet they work together seamlessly to solve the so-called ‘unsolvable triangle’ issues. This integrated approach is why Stratos outperforms other storage networks like Filecoin and AR.
Stratos and AI🔗︎
Yes, it seems like Stratos is truly designed with the future in mind.
Speaking of the future, how does Stratos plan to integrate AI into its ecosystem, and what specific benefits do you envision this bringing to users, developers, and the broader community?
You've already shared a lot about your milestones, but could you elaborate a bit more on your ecosystem and its AI integration? Thank you.
Actually, I began working on AI projects back in 2018, when I was the technical co-founder of an AI company in Montreal. We focused on facial recognition, using cameras to detect people’s bodies and faces to help retailers better understand their customers. At that time, AI was still quite primitive—there were no large models, no OpenAI, no ChatGPT. The technology simply wasn’t advanced enough. I mention this to highlight why AI seemed 'slow' and 'stupid' back then, but today it’s significantly better. There are two key reasons for this shift. First, the development of GPUs—while GPUs and even TPUs were already available in 2018, their performance wasn’t sufficient for training large models. Second, and more importantly, data. AI’s intelligence depends not only on computational power but on having enough data. Without sufficient data, even the most powerful computing resources won’t make AI work effectively.
In 2023, new methods were found for gathering data and using more powerful computation, which led to the development of smarter AI. Many AI companies, like those led by Elon Musk, have advanced significantly, and we’re likely to see even more breakthroughs soon. This is where Stratos comes in—we already provide storage, and we aim to bring computation to the network with Deepseek. We’re also working on delivering a decentralized AI database, which will be crucial for AI systems. Stratos covers the three essential components for AI: storage, computation, and data. This makes Stratos a powerful platform for AI users and developers alike.
We want AI developers to use Stratos to store their data, run computations through Deepseek, and leverage the decentralized database for their vector data. This complete solution will allow developers to pick the components they need—whether it’s storage, computation, or database—similar to building with Legos. If developers need everything, Stratos can provide that as well. That’s our vision, and we’re committed to building this ecosystem for the future.
Stratos Challenges🔗︎
Yes, it definitely seems like a natural fit for Stratos. Now, let's talk about what makes Stratos the key to our success. What has been the most significant challenge you’ve faced while developing Stratos, and how did you overcome it?
Yes, there have been many challenges, for sure. To be honest, over the past four years, we’ve made some missteps. I can give you one example that really cost us a year of development. When we first tried to build the meta-layer—the 'brain' of Stratos, which tracks and retrieves data—we used LevelDB, a database that is commonly used by blockchains like Bitcoin and Ethereum. Initially, we thought it would work well for storing Web3 data and user storage, but after more than six months, we realized it was completely unsuitable. It was too slow and not scalable. As we added more nodes, the performance actually decreased rather than improving. We eventually realized that while LevelDB works for traditional public chains, it wasn’t a good fit for a storage network like ours. This discovery took us another six months, but in January, we finally solved the problem by switching to a better solution. Since then, the network has been improving rapidly, with node sizes growing and the network expanding faster.
Another major challenge we faced was in marketing. We began the project in 2021, and even before that, I knew how critical it was to provide a solid, decentralized solution, especially in the crypto space where people are very focused on authenticity. At that time, the community cared deeply about whether a project was genuinely decentralized or just a facade. However, in today’s market, it feels like many people no longer care about the technology behind the projects. They are more focused on the price action. This shift makes it harder for us to get attention for Stratos because we aren’t offering quick returns like other projects that issue tokens for short-term profit. Unfortunately, people often overlook the value of the technology, and instead, they’re driven by the price of the token.
We understand this challenge, but we are trying a different approach. We want users to try our product and compare it with others. We’re targeting people who care about privacy and self-ownership, especially within the Web3 community. For example, Stratos Drive offers no KYC, no censorship, and ensures that users maintain control of their own data. We're also working with key opinion leaders (KOLs) to try our product, and if they find it valuable, we hope they’ll help us spread the word and improve our marketing efforts. This is the strategy we’re currently pursuing.
Stratos Future Partnerships🔗︎
That’s truly an impressive journey, and quite touching. Looking ahead, can you share any upcoming partnerships or collaborations that could potentially redefine Stratos' role in the industry?
Yes, we’re still in progress, but we are actively engaged with several potential partners. We currently have ongoing discussions with Web2 companies, especially traditional businesses, who tend to be more conservative in their approach. While we are talking with them and even planning pilot projects, we’re not yet in a position to announce anything.
Additionally, we’ve had conversations with some public cloud providers who see the potential in combining decentralized solutions with their centralized infrastructure. They’ve suggested a hybrid approach, mixing both decentralized and centralized networks, and we’re exploring that possibility. However, it’s still too early to share any concrete details.
On the crypto side, we continue to engage with various crypto projects. Whenever we speak to a new project, we encourage them to try Stratos by offering them free traffic and storage space to test our platform. Sometimes, the challenge is that many projects don’t want to invest time learning new SDKs or integrating Stratos into their existing networks. That’s why we believe our upcoming IPFS integration will make it much easier for these projects.
There are already projects using IPFS, and with Stratos, they won’t need to change much—just configure the gateway, and they’ll be able to load their data onto our IPFS system. Although IPFS has free space for end users, it’s not cost-free for business use, but we can help cover those costs during the pilot phase, which could last three to six months, depending on the level of collaboration. In the long term, we aim to reduce their costs, offering a free trial period to attract more partnerships in the crypto space.
For Web2 companies and other traditional businesses, we know that building relationships takes time—we’ll need to meet and have more in-depth discussions to move forward.
Stratos X DeepSeek🔗︎
Yes, it’s exciting to see how creators continue to grow and evolve. I also appreciate the energy in our space—there are so many passionate people raising their hands and eager to contribute. We’ll go through our final question, and then we’ll open the floor to more.
This next question is something our community is particularly curious about: Stratos is known for its potential to support cutting-edge technologies, including Deepseek.
What are the potential benefits and challenges of a collaboration between Stratos and Deepseek in advancing AI technology? This is my question.
That’s a great question. Right now, OpenAI’s ChatGPT and other large models from companies like Facebook, X, and Google are widely used, but there are concerns, especially in sensitive industries like content and business security. When interacting with these AI models, it’s similar to talking to a friend—you have to decide whether you trust them. Businesses don’t want their strategic or sensitive information to be exposed. That’s a major issue with public models.
We aim to build the Deepseek model on Stratos with a focus on privacy. The idea is that everyone will have their own storage, and they can keep their sensitive data secure while still training AI models. For example, if you have personal data like photos and other files, you can train models without risking exposure. You can upload your sensitive data to Stratos storage, and then grant Deepseek permission to use it. However, your personal knowledge or data will stay within your control—stored in your private area, potentially in a decentralized AI database within Stratos, or somewhere within Deepseek’s knowledge base, but only accessible under your permission.
This is an area we are actively researching, and our goal is to address people’s concerns. You can use large models, but you’ll also have peace of mind knowing you’re in control of your data and can trust the model. We’re still working on it, but that’s the direction we’re heading in.
Ending🔗︎
Great! Now, let's move on to the community Q&A session. The floor is open—do any of you have questions for Mr. Bin? This is your chance to dive deeper into this exciting topic. As mentioned in our tweets, we’ll be picking five lucky listeners to receive a light remote miner plus one Zero Door and one ART token each. If you have a question, feel free to ask... though it seems we may be running out of time.
I don’t want to keep Mr. Bin too long, especially when it’s not very safe to keep a top AI entrepreneur like him waiting for free things! But don’t worry, we’ll log all the questions and the Stratos team will follow up with you later.
Mr. Bin, do you have any final thoughts to wrap things up?
Actually, I'm quite familiar with your project, Arkreen. I had the chance to meet the founder in Hong Kong twice—once the year before and again last year. I really appreciate the work you’re doing in the green energy space. I believe there’s potential for a deeper collaboration, especially in terms of storage and computation. It could be an exciting opportunity to leverage clean energy to build a more sustainable network together.
A huge thank you to Mr. Bin and the team for your ongoing support of our Arkreen project. I also know some of our friends from the Stratos community are here today. Arkreen is a deepening network that connects and modernizes decentralized, distributed renewable energy sources to work together with all our partners to create a carbon-neutral Earth. Given Stratos' innovative approach to decentralized AI infrastructure, combined with Arkreen's on-chain registry, I believe we can reduce AI's environmental footprint and establish new Web3 carbon neutrality standards in the future.
Thank you so much for joining us today and sharing such valuable insights, Mr. Bin. It’s clear that Stratos is making significant strides in the DePIN and AI space, and we look forward to seeing how it continues to succeed. Thanks again to everyone for tuning in and being part of Arkreen Consulting. I’m your host for Arkreen Talk today, and we hope to see you next time. Thank you!
Thank you, everyone. Thank you. Sure.