Interview with Julian Peh, CEO and Co-founder of KIP Protocol: Empowering the Future of Decentralized AI

With the rapid development of AI and blockchain technology, decentralized AI (deAI) is becoming an important trend for the future. KIP Protocol, as a pioneer in this field, is providing AI developers with a new platform through its innovative infrastructure and tools, enabling them to deploy and monetize AI products in a Web3 environment. In this interview, Julian Peh, CEO and co-founder of KIP Protocol, shares insights into KIP's vision, technological innovations, and future development directions.

1. What prompted the establishment and development of KIP Protocol? How will it change the landscape of decentralized AI?

Since 2019, we have been actively engaged in commercial operations in the fields of artificial intelligence, machine learning, and big data, providing AI and big data analytics services to globally renowned organizations such as the World Bank.

However, in the past two years, we have observed a clear monopoly forming in the AI field, with a few companies (like OpenAI) leveraging our collective data and knowledge to train their massive models and completely control the regulatory process.

If this situation continues, we will have nothing in an AI-driven future. We are all victims of a large-scale theft of knowledge, and in the future, we will merely be consumers of AI, losing any rights to economic participation. Ironically, these massive models are trained using our data.

We are acutely aware that we must fight for the decentralization of AI. If we are to have a voice in an AI-driven future, everyone needs to regain ownership. This is the driving force behind our progress. We believe the current path of AI development is wrong, and such a future is not one that my team and I are willing to see.

Therefore, when we discovered the lack of infrastructure and tools serving AI developers to create decentralized AI products, we built KIP Protocol as a foundation to help AI developers deploy and monetize their work on Web3. To date, we have deployed large projects on KIP Protocol, and our infrastructure is flexible enough to allow independent developers to join.

We hope to attract a large number of AI developers to Web3 to address their fundamental issues in deployment, connectivity, and monetization. This will catalyze the development of the entire decentralized AI ecosystem.

2. Many decentralized AI projects focus on decentralized inference (or models) or AI agents. Why does KIP Protocol focus on the deployment and monetization of decentralized AI?

The ultimate goal of KIP Protocol is to bring a large number of AI end-users into Web3. To achieve this goal, we first need many excellent AI products to attract and serve users in Web3.

This also explains why KIP Protocol focuses on three categories: AI applications, models, and data, as a usable AI product is composed of these three components.

KIP Protocol focuses on solving the issues of how to quickly deploy AI products and achieve monetization in Web3. If the monetization issue is not addressed, digital ownership is incomplete, and decentralized AI will only be a niche field limited to true believers.

There are far more AI developers in the world than blockchain developers. If KIP Protocol can solve the monetization problem for decentralized AI, we will attract a large number of AI product developers into Web3, who will be responsible for attracting users.

Our goal is to establish a vibrant, revenue-driven decentralized AI business ecosystem. Only in this way can decentralized AI reach its first stable state and thrive.

3. Please elaborate on these three types of AI value creators: AI applications, models, and data, and how they reflect KIP Protocol's unique approach to decentralized AI?

To build an AI product ready for users, it is necessary to link AI models with AI applications (for example, a chatbot that passes end-user instructions to the model). Depending on the nature of the product, external datasets can also be integrated into this setup. Therefore, an AI product is composed of AI models, AI applications, and datasets.

One of our unique insights is that in a fair and competitive AI ecosystem, models, applications, and datasets are often not owned and developed by the same group of people or companies. The people who create models are often different from those who create applications, and those who own large datasets are also different, as their skills and business motivations are entirely distinct.

Thus, KIP Protocol's vision for decentralized AI is to build an ecosystem where even if models, applications, and data are owned by different owners, they can collaborate in a permissionless manner to provide useful and diverse AI products to end-users and be able to earn and share revenue from users.

4. What technical challenges do AI developers face in the monetization of decentralized AI, and how does KIP Protocol address these issues?

When AI developers wish to deploy fully functional decentralized AI products, they can use KIP Protocol's contracts and infrastructure as a complete toolkit to address the following foundational issues:

1. Deployment issues: Easily tokenize and gate AI assets (such as AI applications, models, and datasets).
2. Connectivity issues: Ensure data flows between models, applications, and assets, even if they are owned by different parties.
3. Monetization issues: Ensure the recording, accounting, and settlement of interactions between AI assets, even if they are owned by different parties.

By addressing these three issues, KIP Protocol effectively makes it possible to create a commercial ecosystem in decentralized AI, as developers can not only deploy products but also ensure they are compensated.

Our tech stack has five notable components:

1. Token gating that optimizes gas fees: Achieving comprehensive connectivity and functionality of on-chain AI assets through trustless NFT and SFT-supported selective access while minimizing on-chain costs.
2. On-chain micropayment infrastructure: Seamlessly trading inference and data with AI components owned by other ecosystem participants and being able to settle these transactions in a permissionless manner.
3. Account abstraction: Providing a Web2 user experience for decentralized Web3 AI interactions.
4. Semi-fungible tokens: The optional ERC 3525 standard allows for the fractionalization of AI asset ownership, enabling flexible financial structures to capitalize and monetize AI assets.
5. Privacy/security: Supporting zero-knowledge proofs.

The early version of KIP infrastructure won the Tencent Cloud Award at the 2023 Chainlink Hackathon. As experienced AI developers, we essentially built the infrastructure and tools we needed but could not find.

5. What mechanisms within KIP Protocol facilitate automated revenue sharing among app developers, model creators, and data owners, and how is transparency maintained in these transactions?

Whenever AI developers wish to deploy AI assets (whether models, applications, or datasets) on-chain, they set the price for each query during the deployment process, which is recorded in the token metadata. Whenever a user wishes to interact with an application and connect with the model and/or dataset associated with that application, they pay the total query price for all three parts. Therefore, if the application charges 0.1 $KIP per query, the model charges 0.2 $KIP, and the dataset charges 0.3 $KIP, the total price paid by the user will be 0.6 $KIP per query. This amount is then automatically distributed to each contributor through KIP Protocol's contracts.

This approach allows everyone to set their own prices while letting market forces operate freely. For example, if you charge 10 $KIP per query for access to your dataset, as long as your data is indeed rare and high quality, you might find customers. However, if someone offers a better dataset for just 5 $KIP, you may need to adjust your price if you want to compete.

In this way, KIP Protocol ensures fair distribution of revenue and achieves transparent and automated revenue sharing through smart contracts.

6. KIP recently announced the latest developments in its collaboration with Open Campus. Why is KIP so focused on the education sector as an infrastructure provider?

We are collaborating with Animoca's Open Campus to launch Open Campus U, an AI-driven education platform that will serve professors, universities, and college students.

Education is an $8 trillion market involving hundreds of millions of people, and the application of AI in education is very evident. In fact, we have been providing AI services to universities since 2019, so we are very familiar with this field.

Open Campus and KIP Protocol are completely aligned in our goals, opening new opportunities for educators to thrive in the digital economy by integrating KIP's cutting-edge AI technology to meet the ever-changing needs of global learners and educators.

We aim to enhance the quality of education by making the educational experience more adaptive and personalized. Blockchain technology can empower educators by allowing them to transform their knowledge into assets that can be owned, traded, or monetized.

We are very excited about these opportunities and will invest significant resources to deploy KIP through Open Campus U in the education sector. More content is yet to be announced, so stay tuned.

7. Can you discuss the experimental aspects of the protocol, such as decentralized human feedback reinforcement learning (RLHF) and inspection nodes, and what potential impacts these innovations may have on the broader AI and blockchain ecosystem?

KIP is currently ready to serve AI product developers and attract users into Web3. However, we are still constantly thinking about how to get more end-users directly involved in the decentralized AI ecosystem. We need to raise awareness of the concept of digital property rights in AI, and the best way to do this is to encourage actual participation.

One idea we are experimenting with is to incentivize end-users to participate in the decentralized AI ecosystem through activities like RLHF or model evaluation. The goal is to fairly reward all users for their economic participation in the AI ecosystem, making them capital owners in an AI-driven future.

More participants, whether product developers or end-users, will help strengthen the robustness of the decentralized AI ecosystem built by KIP.

8. What recent developments can you share?

The next 4 to 8 weeks will be filled with news related to product launches and community activities. KIP will soon bring a leading entertainment AI platform with a massive user base into Web3. This is very exciting and will significantly increase our actual user base.

Based on our collaboration with Open Campus, we will also establish deep cooperation with another large Web3 project, becoming their official AI partner to propose AI application cases for their platform.

Finally, KIP previously released a series of NFTs, and we will launch a long-awaited plan to integrate these NFTs into a broader long-term membership program, codenamed "KIPVERSE."

The coming weeks will be very busy for KIP, so please stay tuned!

About KIP Protocol

KIP Protocol builds a Web3 underlying protocol for AI app developers, model creators, and data owners, enabling AI assets to be easily deployed and monetized while retaining full digital ownership.

KIP aims to establish a new AI business ecosystem to address the challenges and issues faced in the deployment of decentralized AI, ensuring that everyone can enjoy the economic benefits brought by AI.

The KIP team consists of experienced PhDs and technical experts dedicated to AI research since 2019, who also have a strong professional background and rich experience in the Web3 field, committed to promoting the decentralization of AI and becoming a catalyst for the decentralized AI wave.

To learn more, please join the official Chinese community: https://t.me/KIPProtocol_CN