How AI converges with Web3: Interview with Footprint Analytics CEO

Footprint Analytics is dedicated to creating a structured data platform that bridges the gap between Web2 and Web3 data.

We specialize in structuring data. Despite the relative advantage of Web3 over Web2 in transparent on-chain data, certain challenges remain. These include the nascent status of the industry, a lack of standardized practices, and a lack of organized data. As a result, data application becomes problematic.

To illustrate, consider the scenario where you want to access transaction data on Opensea from multiple chains such as Ethereum, Solana, and Polygon. This process involves understanding OpenSea’s business model, studying smart contract code, and sequentially extracting transaction data from each chain.

This process is complicated. First and foremost, it’s complicated and prone to errors throughout the data collection process. Second, it is technically complex, given the differences in ledger design and data structures across chains. Finally, it leads to a waste of resources. In a scenario where 1,000 people need this data, they’d have to go through a similarly complex process 1,000 times. This significant repetition significantly hinders data collection efficiency and wastes computing resources.

This brings us to the purpose of Footprint Analytics: to abstract data from disparate sectors such as GameFi, NFTs, and DeFi and establish standardized data practices for the Web3 industry. This, in turn, will enable developers and industry participants to access and analyze data efficiently and accurately.

To date, we’ve launched platforms on more than 20 blockchains, organized into three core segments:

1. Footprint Growth Analytics as an Industry Solution: Tailored solutions for Web3 projects in marketing growth and operational analytics, similar to a Web3 version of Google Analytics, driving projects towards data-driven growth.
2. Zero-Code Data Analysis Tools: Providing an experience similar to ChatGPT, this tool allows users to obtain data analysis reports through simple queries and responses. In the foreseeable future, the use of on-chain data will be greatly simplified – no complicated understanding of Web3 business logic or advanced programming skills will be required, streamlining the transition from Web2 to Web3.
3. Free Unified API: Through a unified multi-chain and cross-chain API, this feature facilitates cross-chain data access across multiple chains, providing users with a seamless experience to retrieve data from multiple chains at no cost.

As a data platform, Footprint is a natural fit with AI. AI encompasses three key facets: computing power, data, and algorithms. Among these, computing power is the foundation that underpins AI model training and execution. At the same time, data is the essence of AI, and algorithms dictate AI performance, including model accuracy and application effectiveness.

Of these, data is undoubtedly the most important and indispensable. Data is the lifeblood of industries and projects, and its importance extends to key areas such as privacy and compliance, where its value is immeasurable. Data may be beyond purchase, given its involvement in privacy and compliance issues. AI acts as both a consumer and a producer of data.

Currently, Footprint’s application of the convergence of data and AI encompasses several primary aspects:

During the data content generation phase, the contribution of AI within our platform is critical. Initially, we use AI to generate data processing code, providing users with a more streamlined data analysis experience.

More specifically, we are driving innovation in two specific directions.

First, we are curating and categorizing reference data. Taking recently deployed contracts on the blockchain as an example, our AI can autonomously determine the protocol to which a contract belongs, the type of contract, and even whether the contract falls under categories such as LP or Swap on Dex platforms. This intelligent structuring and classification greatly improves data accessibility.

Second, we can generate higher-level domain data based on our reference data. For example, we use AI to create data within domains such as GameFi, NFT, etc., providing users with richer data resources. This approach enhances the quality of data content and enables users to better understand data across different industries.

To improve the front-end user experience, we have introduced an AI-based intelligent analysis function. As mentioned above, when users engage Footprint for data analysis, they encounter an experience similar to a conversation with ChatGPT. Users can ask questions and immediately receive corresponding data analysis reports. The underlying logic involves translating text into SQL queries, dramatically lowering the entry barrier for data analysis.

Finally, when it comes to user support, we’ve developed an AI-powered customer service bot. We feed AI with data from Footprint, which spans GameFi, NFT, DeFi, and other areas, to build a custom AI customer service bot for Footprint. This AI bot provides immediate assistance to users by answering questions related to the use of Footprint, including data types, data definitions, API usage, etc. This greatly increases the efficiency of customer support while reducing the amount of manual work.

However, it’s worth noting that while AI applications can increase productivity and help solve most challenges, they may not be omniscient. Based on our data processing experience, AI can assist in solving approximately 70% to 80% of challenges.

From a broader perspective, regardless of the domain in which AI is applied, a critical consideration is the level of acceptance of AI’s fault tolerance. Different application scenarios have different fault tolerance requirements. There’s a need to balance the accuracy and reliability of AI against people’s tolerance for error.

For instance, in healthcare, the decision to trust either AI or a physician may involve trust-related challenges. In the investment space, AI can provide factors that influence the direction of BTC prices, but people may still have doubts when making actual buy or sell decisions.

However, precise accuracy may not be paramount in marketing and operational analytics, such as user profiling and tiering, because minor errors won’t significantly impact. As a result, error tolerance is more readily accepted in these contexts.

Currently, Footprint is primarily focused on data in its efforts to integrate AI with Web3, which presents its own set of challenges:

First, the first challenge is data generation, specifically providing high-quality data for AI to achieve more efficient and accurate data generation capabilities. This relationship between AI and data can be compared to the engine and fuel of a car, where AI is the engine and data is the fuel. No matter how advanced the engine, a lack of quality fuel will prevent optimal performance.

This raises the question of how to generate high-quality data, for example, how to quickly and automatically generate data in areas such as GameFi, NFTs, DeFi, and others. This includes automatically organizing the data connections, essentially creating a data graph. More specifically, it involves determining factors such as the protocols to which contracts are associated, the types of contracts, the providers, and other pertinent details. The main goal of this process is to consistently provide the AI with high-quality data to improve its efficiency and accuracy in data production, thus creating a virtuous cycle.

The second challenge is data privacy. While Web3 is fundamentally committed to decentralization and transparency, the need for privacy may become paramount as the industry evolves. This includes protecting users’ identities, assets, and transaction information. This situation presents a dilemma: the transparency of data on the blockchain gradually decreases, limiting the amount of data accessible to AI. However, this issue will be addressed as the industry progresses, and homomorphic cryptography is a possible solution.

In conclusion, the convergence of AI and Web3 is inherently intertwined with a core problem: data accessibility. In essence, the ultimate challenge for AI lies in its access to high-quality data.

I believe achieving significant integration and adoption of Web3 and AI depends on addressing two fundamental challenges. First, there’s a need to provide enhanced services to Web3 builders and developers, especially in areas such as GameFi, NFTs, and social platforms. Second, it’s imperative to reduce the barriers on the application front to ensure a smoother user entry into the Web3 landscape.

Let’s start with serving the developer community. In this area, two primary types of applications stand out.

One category is AI-powered development platforms. These platforms use AI technology to automate the creation of code templates. Whether for building DEX platforms or NFT marketplaces, these platforms can intelligently generate code templates tailored to the specific needs of developers, significantly increasing development efficiency.

In games, AI can speed up the creation of game models and the generation of images, thus accelerating the game development and launch process. These platforms have allowed developers to focus more on creativity and innovation rather than excessive time on repetitive, basic tasks.

The other category revolves around AI-powered data platforms. These platforms use AI to autonomously generate domain-specific data in various industries such as GameFi, NFTs, SocialFi, and DeFi. The goal is to lower the threshold for developers to use and apply data, and simplify data analysis and use.

Through AI, these platforms can automatically generate diverse data sets, enriching developers with rich data resources and improving their understanding of market trends, user behavior, and more. By providing developers with comprehensive data support, these data platforms remove data utilization barriers and catalyze inventive applications’ emergence.

Mass adoption has always been a key challenge in the Web3 space. For example, the market has recently seen the emergence of blockchain solutions with virtually negligible fees aimed at increasing transactions per second (TPS). In addition, solutions such as the MPC wallet effectively address the primary barrier to migration from Web2 to Web3 by overcoming migration challenges.

The solution to these challenges doesn’t depend solely on AI technology but is intertwined with the holistic evolution and development of the Web3 ecosystem. While AI plays a key role in improving efficiency and reducing barriers, the underlying infrastructure and growth of Web3 remain key factors in solving the mass adoption problem.