Interview with GAIB Founder Kony: Breaking the "Capital Dilemma" of AI Infrastructure, How GAIB Turns GPUs and Bots into Income-generating Assets in the DeFi World?

Interview: The Round Trip

Compiled & Organized by: Yuliya, PANews

In the increasingly capital-intensive competition of the AI supply chain, physical assets such as GPUs and robots are becoming the world's most scarce and valuable production factors. With GAIB announcing that it will conduct TGE on November 19, this RWA × AI × DeFi model has also attracted greater external attention.

In the new series Founder’s Talk of “The Round Trip” co-produced by PANews and Web3.com Ventures, host John Scianna and Cassidy Huang invite Kony Kwong, co-founder and CEO of GAIB, to delve into the story behind the founding of GAIB, how to design a robust risk control model, and the complete business path of turning computing power into tradable assets.

*Note: This video interview was conducted on October 30, and some data may differ from current figures or dynamics.

The establishment of GAIB - Bridging the gap between AI and finance

PANews: First, please introduce yourself and the opportunity that led to the founding of GAIB. We know that you were a venture capitalist at L2 Iterative Ventures (“L2IV”) and have a traditional finance background. What experience made you realize this unique problem existing in the current AI field?

Kony: Before entering the VC field, I worked in traditional finance, including credit research, equity research, and traditional investment banking. Later, I developed a strong interest in cryptocurrencies, which led me to join a large exchange, where I was responsible for its overseas expansion, leading mergers and acquisitions with other exchanges, wallet companies, e-sports platforms, and more.

Later on, my partner and I established a cryptocurrency VC fund focused on investments in the infrastructure sector, such as projects like ZK, Layer 1, and cross-chain MEV.

The real turning point that motivated me to establish GAIB was at the end of 2023 to the beginning of 2024. At that time, the crossover between AI and Crypto began to emerge, and everyone was exploring the possibilities here. As an AI enthusiast, I conducted in-depth research on this. Since the release of ChatGPT-3.5, I have invested a lot of energy, and even built an AI agent myself. At that time, there were almost no mature frameworks available, and everything had to start from scratch, such as how to perform data vectorization, how to search, how to implement RAG (retrieval-augmented generation), and how to establish long-term memory (the context window was very short at that time).

I found that these two fields I like the most are merging, so I started to delve into the Crypto × AI track. But at that time, most companies in the market were doing:

• Decentralized computing power market (such as Akash and Render);
• Decentralized Data Annotation (using blockchain as an incentive mechanism to encourage public participation in data collection);
• Some also focus on AI agents.

But what surprises me is that almost no companies approach it from a “financial” perspective. This is unusual; as someone with a financial background, I am well aware that the most important thing in the early stages of an industry is to solve the “capital problem”. The entire AI industry, especially in terms of AI infrastructure, is experiencing exponential growth in computing power demand, which directly leads to an exponential demand for computing assets like GPUs. This niche is extremely capital intensive, building data centers, purchasing GPUs, and deploying infrastructure is extremely costly.

At that moment, I thought, why don't we do something here? For me, the core spirit of blockchain lies in creating new markets and new types of assets, which is exactly what DeFi and “DeFi Summer” have inspired us to do. Hence, I came up with the idea of providing financial services for AI infrastructure, as this part captures most of the value in the AI supply chain.

I shared this idea with my co-founder Alex. His background is more aligned with the semiconductor and AI fields; his family runs one of the top seven chip manufacturers in the world—Realtek. He also operates a cloud service company named GMI Cloud. He personally feels all the pain points I mentioned: As a startup cloud service company, acquiring capital is extremely difficult. The reasons are twofold:

• Firstly, GPUs and computing power were still a very new asset class at that time (a year or two ago), and no one truly understood it;
• Second, cloud services themselves are also very new.

So, Alex and I hit it off immediately. We believe that a financial company focusing on the AI field should be established. This has happened in all other industries, whether it is coal, real estate, or any other industry. Thus, we decided to establish GAIB last year.

The return on investment in AI infrastructure and the business model of GAIB

PANews: Indeed, the AI field requires huge capital expenditures, and the return on investment (ROI) can take several years. What are your thoughts on the investment returns for AI infrastructure like GPUs? Meanwhile, how does GAIB collaborate with cloud service companies and help them shorten their capital turnover cycle?

Kony: In fact, most people may not understand that GPUs, as an infrastructure, have quite considerable profitability.

• First of all, large AI companies or gaming companies require a substantial amount of GPUs and typically enter into long-term contracts of 2-3 years with cloud service providers to ensure stable computing power. This brings robust cash flow to cloud service providers.
• Secondly, the AI computing power used on demand is extremely scarce. For example, a few months ago, when ChatGPT launched a new feature that could convert images into specific styles, it led to OpenAI itself experiencing a shortage of computing power, indicating that the market is still in a clear state of supply and demand imbalance.
• More importantly, the typical return on investment cycle for GPUs is actually 12 to 18 months. This means that, from a revenue perspective, the annualized return on investment (ROI) can reach 50% to 100%, which is exceptional compared to other asset classes.

Regarding the second question, who we are collaborating with. My co-founder Alex's company GMI Cloud is naturally one of our first partners, and it is the starting point of our project. However, in the nine months following that, due to the exponential growth of the market and the surge in demand, we received a large number of collaboration requests and established a strong project reserve. Currently, we are collaborating with over 10 global “Neo Cloud” providers and Nvidia cloud partners, with business coverage in Thailand, Taiwan, Singapore, Hong Kong, Japan in Asia, the United States, Canada in North America, and Norway, Iceland, and Denmark in Europe.

We tend to collaborate with companies that are typically Nvidia cloud partners. For audiences who are not familiar, this means these companies have passed Nvidia's review and have the license and capability to provide software and hardware services. More importantly, they receive official recommendations and preferential treatment from Nvidia in acquiring the latest GPUs and quality clients. Therefore, most of our partners are Nvidia cloud partners.

This leads to our unique advantages in this field:

• First, our founding team is itself an operator in this field, and we have a deep understanding of its economic model and GPU business.
• Second, we have a network of cooperative clients.
• Thirdly, the capital solutions we offer are more flexible, better meeting their needs in terms of underwriting time and duration.

As a result, our collaborative project reserves have been continuously growing.

How does GAIB guarantee returns and control risks?

PANews: You mentioned collaborating with cloud service providers around the world, but the cost structures for electricity, data centers, and other factors vary greatly by region. How do you ensure a relatively consistent return on investment (ROI) among different partners? Additionally, you also mentioned collaborating with Nvidia cloud partners. Does this mean you have a set of standards to ensure the credibility and operational capabilities of your partners?

Kony: Yes, it is very difficult to ensure that the ROI is completely consistent across different cloud service providers because they each have different fee and cost structures. For example, the costs of electricity, facilities, and data centers in Asia are completely different from those in the United States.

Therefore, when trading with these cloud service providers, our focus is not on cost, but on net cash flow recovery. We will evaluate:

• If we give the other party 100 dollars, when and how will they return 100 dollars to us.
• All factors such as electricity costs, facility costs, and even depreciation costs that may affect asset value.

As for the trading structure, it depends on the specific agreement we reach with them. Sometimes we adopt a fixed-rate model, requiring a fixed annual return rate. In other cases, we prefer to directly invest in the assets themselves and then take a certain percentage of the total income generated, such as 50% to 70%. In this model, our protection is stronger.

This comes down to the actual structural design of the transaction and our experience. We can request various protective clauses, such as prioritizing the recovery of our total investment and returns before the other party's company distributes any profits. In summary, we will set up various protective mechanisms to ensure that our funds are repaid first.

In addition, we have two hard criteria when trading with these cloud companies:

• There must be real assets as collateral, and they must be over-collateralized. When we provide funding, there must be physical assets as support. More importantly, we require over-collateralization. For example, if the counterpart only has a GPU worth $100, we might only provide $70 to $80 in funding, which means we have an over-collateralization rate of at least 1.3 to 1.5 times. This way, in case of any issues, even if the asset is sold at a discount, our principal still has a fairly safe buffer.
• The company must have signed clients and a good payment record. This ensures that the GPUs we invest in are not sitting idle burning money, but are being effectively utilized and can continuously generate cash flow to repay the funds we provided.

If a company does not meet either of these two standards, we will not continue to advance the transaction negotiations.

From “spices” to tokenization, the core concept of GAIB.

PANews: This risk control model sounds very robust. Additionally, we are very interested in the name GAIB, which seems to be related to the famous science fiction novel “Dune.” Could you explain the origin of this name and share your thoughts on the position of GPUs in the AI value chain?

Kony: Yes, we are all fans of Dune, and the name GAIB is indeed inspired by Dune. In fact, it is also an acronym for GPU, AI, and Blockchain. You could say we are a “Global AI Infrastructure Blockchain” platform.

This analogy is very apt. In the universe of “Dune,” the “Spice” is the most precious and important commodity. Similarly, in our current AI era, computing power is everything. Whether you are using ChatGPT, Claude, or Perplexity, the core fundamental unit you are discussing is computing power. Therefore, the role of computing power is very similar to that of the “Spice.”

As for the position of computing power in the AI supply chain, I like to describe it using a “smile curve.” This means that value is primarily concentrated at both ends of the curve.

• One end is the application layer, as they hold the pricing power and users. However, at present, the curve of this market is not steep enough, as most applications only started to become truly profitable this year, having previously been burning money without forming large-scale commercial applications. But we are seeing more and more applications emerging.
• The other end, which is the left side of the curve, is the AI infrastructure, including GPUs, data centers, and even robotics manufacturing companies.

Regardless of how the application layer evolves, it will rely on the core AI infrastructure. As I mentioned before, whether you are using the ChatGPT model or the Claude model, it ultimately depends on the underlying GPU chips to provide the power.

I like to use the “Visa card” analogy: no matter which bank issues your Visa card, every time you make a transaction, Visa earns a little money from it. The same goes for GPUs; whenever you invoke any model or use any AI application, the GPU is running, providing computing power, and generating revenue from it. This is why we focus on core AI infrastructure, as this segment has enormous expansion potential with the continuous growth of the application side.

How does GAIB turn GPU/computing power into on-chain assets?

PANews: Financializing AI infrastructure sounds like a great entry point. So, what are the next steps? How do you plan to build a complete financial stack on this foundation?

Kony: This is a very good question. Within GAIB, we refer to ourselves as an “economy” because we are the bridge connecting off-chain RWAs with on-chain DeFi economies. The process we use to handle these assets is mainly divided into three steps:

• Asset Digitization: We must first convert these physical assets into digital form. Otherwise, their data, asset value, and other information cannot be represented and utilized on the chain.
• Asset Tokenization: After assets are put on the blockchain, the next step is to convert them into useful financial instruments or products. For example, can they generate returns? Can they be used as collateral for lending? Based on this, we have developed different products.
• Injecting Liquidity: Once these assets are on-chain, if they have no utility or trading channels, they are essentially just a pile of useless data. Therefore, we have been expanding the use cases for these assets, including integration with lending protocols, DEX, derivatives protocols, etc., to truly incorporate them into the on-chain economy and create a closed loop.

These are the three things we are doing. With this core infrastructure, we can handle any type of AI infrastructure asset. We started with computing power and have proven that this path is feasible. Currently, we have successfully tokenized assets worth approximately $30 million on-chain.

Now, we are preparing to expand into what we see as the next big trend—robots. If you consider AI to be the “brain,” then robots are the “body” that interacts with the physical world. Similar to GPUs, robots have physical hardware and are about to undergo a significant transformation in their profit models. Future robotic models will be completely different from the large mechanical arms in traditional manufacturing and will become more consumer-grade.

We recently announced a partnership with a Nasdaq-listed company called Primech, which primarily produces cleaning robots. We are exploring the tokenization of these robots, as they have adopted a new business model that we refer to as “Robot-as-a-Service (RaaS).” Under this model, we have both hardware assets and can generate stable monthly revenue, which is the perfect model for us to create a product that provides users with stable AI-related returns.

AI Dollar, GAIB Token and Ecosystem Outlook

PANews: Sounds very exciting. You mentioned the integration with lending protocols, and the integration of DEX is relatively permissionless and easier to implement. But how are you advancing in the lending market? Can you disclose some specific collaboration agreements?

Kony: In the lending market, we are about to integrate with Morpho and many other similar protocols. Additionally, there are various types of lending protocols available on different blockchains for us to utilize. For example, on Plume Chain, there is a lending protocol specifically designed for RWA, so we are working hard to integrate with as many blockchains as possible to allow our assets to be utilized as widely as possible.

PANews: Last month, there were some cases in the NFT field where creative strategies were used to release the liquidity of “non-liquid assets.” I was wondering if someone could create an “AI strategy,” which utilizes these tokenized AI assets to profit from trading fees, and then reinvest the earnings into more AI infrastructure?

Kony: This is an interesting idea. This is also one of the reasons why we are launching the AI-supported stablecoin or synthetic dollar – we call it “AI Dollar”. The goal of launching the AI Dollar is to make it a universal “umbrella” covering all assets on our platform. The value of the AI Dollar will be supported by all the different types of tokenized AI infrastructure assets that we introduce, including computing power and robots.

In this way, users have a unified unit that they can easily use to generate returns, and it can be integrated into any DeFi protocol they desire. Therefore, AI Dollar is a single entry point we provide for users to access the entire world of AI infrastructure.

PANews: How can users earn profits through AI Dollar? Do they need to stake on your platform?

Kony: Yes, it's similar to other models. For AI Dollar, you can stake it to get a staking certificate version. This staking certificate will continuously generate profits from the underlying computing power and robotic assets.

PANews: So, what is the vision and role of GAIB's native token in the entire ecosystem?

Kony: The GAIB token is a crucial element in our entire ecosystem. It is not just an ordinary governance token; it has real utility.

As I mentioned before, GAIB is an infrastructure platform. One of the core components of this infrastructure we are building is our node network, which we call the “verification network” or “node coordination network.” This network requires all tokenized GPUs to continuously run a node and report data to our network to ensure that these assets genuinely exist, operate normally, and generate returns.

To ensure the security of this network, we require users to stake our GAIB tokens. We have implemented mechanisms of some re-staking protocols. This means that the GAIB tokens provide economic security for the network we offer.

Secondly, the GAIB token is of course the core of all incentive measures in our ecosystem. Whether it is additional earnings, extra incentives, extra rewards, or DeFi integration activities, all behaviors we encourage will be driven by the GAIB token.

Therefore, the GAIB token is at the core of GAIB operations not only at the technical infrastructure level but also at the governance and incentive levels.

PANews: Finally, we see many decentralized computing power providers in the market, such as Io.net and Akash, but they seem to be more focused on Web3 cloud infrastructure. Do you think GAIB, which focuses on serving companies in the Web2 market, will intersect with these Web3 projects in the future?

Kony: I believe their reasons for existence are different. Decentralized computing markets like Akash or Io.net are intended to serve as aggregators, gathering various idle resources, whether consumer-grade GPUs or enterprise-grade GPUs, and providing users with a unified API to access this computing power.

This model is indeed suitable for certain users, as the costs may be lower for small-scale deployments or small-scale use cases. However, if you need to conduct large-scale deployments, such as requiring thousands of GPUs to train a large model or provide production-level services, you may still need to have conversations with those large traditional cloud companies or emerging cloud companies that we are collaborating with.

So I believe that the market is large enough to accommodate their respective niche products and services.