The journey of building infrastructure and applications for Web3 AI is a long marathon.

The trustworthy verification and computing framework of AI is the core area that the web3AI infrastructure needs to tackle.

Written by: Haotian

Recently, I had conversations with many developers at the forefront of web3AI Build and discovered that working around web3AI infra is much more complex than I had imagined.

1. Currently, most actively web3 AI projects are generally meme-ified, boasting a lot of stories that are impossible to realize and implement. The key is that they attracted most of the attention and liquidity by quickly issuing tokens to enter the market, along with the mess left behind after the short-term bubble burst (negative EV). This is mainly due to the overly appealing narrative of AI + Crypto, while the challenges of its actual implementation are too great, making it naturally a heavy disaster area for token issuance based on narrative from the very beginning;

2. The web3AI infrastructure is essentially a reconstruction of the web2 AI infrastructure, which is often a thankless task. It's similar to how Crypto initially challenged centralization in the name of decentralization; for a long time, building decentralized network architectures was criticized as redundant and meaningless, until subsequent DeFi applications found some value capture points.

The current dilemma of web3AI is no different from the initial vision of decentralized Crypto. Most people are still used to casually saying "What’s the use of web3AI?" But let’s not forget that decentralized computing power aggregation, distributed inference, and distributed data annotation networks can all find entry points in terms of training costs, performance, and practicality. It can only be said that the road ahead is long and fraught with obstacles, but its significance is profound.

3. The cost of building and expanding the web3AI infrastructure during the trial and error period is relatively high, requiring strong rationalist support. For example, it is well known that web3AI requires the construction of a data layer, but cleaning up the massive on-chain and off-chain data requires a significant amount of server operation and maintenance as well as development costs. At the same time, the costs of integrating mature web3AI APIs, computing power, algorithm tuning, etc., also require investment. If these costs are focused on Agent applications, it is possible to quickly explore commercial monetization models. However, if the focus is on the infrastructure level, in the current market context where the technology narrative is not very popular, it poses a challenge for many developer teams.

What's more troublesome is that, unlike traditional web2 infrastructure, web3 AI also needs to solve the coordination problem of off-chain data and on-chain verification, the model distribution and update mechanism under the P2P network, and the complex design of replacing traditional business models with Tokenomics incentives. However, the short-sightedness of capital and the speculative atmosphere of market preference have caused some hot money to flow into Agent applications that are rushed online purely for the sake of hot spots, making it difficult for teams that are really working at the infrastructure layer to get enough support.

4. The illusion problem of large models compatible with the "black box" property in web3AI infrastructure poses significant challenges to their security and reliability in specific scenarios. See @SlowMist_Team.

Recently, the output regarding the MCP security vulnerabilities suggests that the professional security audits around MCP are already capable of supporting Slow Fog's future positioning as an AI auditing company. This is just a concrete case that verifies the various unknown security challenges faced by AI LLMs as foundational data sources integrated into web3 AI infrastructure. However, the issues surrounding web3 AI infrastructure are far from limited to these; there are also directions focused on constructing verifiable computing frameworks through web3 cryptographic verification and on-chain consensus mechanisms to ensure that the AI inference process can be traced and verified, among other aspects.

In fact, the trustworthy verification and computation framework of AI is the core area that web3AI infra needs to tackle. Currently, large models have a significantly limited adoption rate in professional fields such as finance, healthcare, and law when handling highly sensitive information due to the inability to provide verifiability of the reasoning process. The maturity of web3 AI infra, such as zkVM at the base layer, decentralized Oracle networks, decentralized Memory solutions, etc., can provide AI with a verifiable and provable computation framework, fundamentally helping AI achieve rapid expansion in vertical scenarios.

That's all.

The journey of building infrastructure and applications for web3AI will not happen overnight; it is a long marathon. Those who can truly build infrastructure and application ecosystems that solve real-world problems, who can balance hype and value in the Go-To-Market process, and who can find practical business loops while maintaining technological foresight, will be the ones who truly emerge victorious in the industry.