Deciphering the Expanding AI Ecosystem of Web3

• The Evolution of Generative AI and Its Impact on the Blockchain Space
• The Harmonious Intersection of AI and Blockchain
• Web3's AI Tech Stack: The Infrastructure Layer
• Middleware Layer
• Developer Tooling & Application Hubs
• Application Layer
• Investor Outlook

The Evolution of Generative AI and Its Impact on the Blockchain Space

In just over a year since the introduction of ChatGPT, the world has witnessed how generative AI has arguably become a central narrative in technological conversations of today. Prompted largely by the early successes of OpenAI, investor interest in large language models (LLMs) and AI applications has spiked dramatically. As a testament to this growing trend, an impressive $25 billion in funding was raised in 2023, which indicates a 5x Year-on-Year increase. This is largely due to the potential multi-trillion-dollar market opportunity that investors see in this emerging field.

The Harmonious Intersection of AI and Blockchain

There is a strong synergy between AI and blockchain technologies, which explains why there is a blossoming AI ecosystem within Web3. However, the intricacies of these protocols, the actual vs hyped potential, and their collective functioning have been a source of confusion. This has led to the need for a detailed exploration and mapping of the Web3 AI supply chain, the various layers in the tech stack, and the competitive landscapes that are emerging. By the end of this discourse, there should be a basic understanding of how the ecosystem operates and what to anticipate next.

Web3's AI Tech Stack: The Infrastructure Layer

Generative AI is powered by LLMs which in turn are reliant on high-performance GPUs. The three key tasks performed by these LLMs are training (model creation), fine-tuning (sector/topic specialization), and inference (executing the model). This layer can be subdivided into general-purpose GPU, ML-specific GPU, and GPU aggregators. Each of these has different capabilities and use-cases. Peer-to-peer markets are incentivized by crypto to endorse secure decentralization. It's critical to note that the actual GPU processing occurs off-chain.

Middleware Layer

Although the previous layer allows permissionless access to GPUs, a middleware layer is necessary to connect this computing resource to on-chain smart contracts in a trust-minimized manner. This is where zero-knowledge proofs (ZKPs) come into play. ZKPs are a cryptographic method that allows one party to prove to another that a particular statement is true without revealing any additional information.

Developer Tooling & Application Hubs

Beyond ZKPs, Web3 developers need tooling, software development kits (SDKs), and services to efficiently build applications like AI agents and AI-powered automated trading strategies. Many of these protocols also serve as application hubs, where users can directly access finished applications that were built on their platforms.

Application Layer

Finally, at the top of the tech stack, there are user-facing applications that utilize Web3's permissionless AI processing power to complete specific tasks for a variety of use-cases. This market segment is still in its infancy and relies on centralized infrastructure. However, early examples include smart contract auditing, blockchain-specific chatbots, metaverse gaming, image generation, and trading and risk-management platforms.

Investor Outlook

While the whole AI tech stack holds promise, the current investment opportunities seem more promising in infrastructure and middleware protocols. This is because of the uncertainties around how AI functionality will evolve over time. Regardless of how it evolves, Web3 AI applications will undoubtedly require massive GPU power, ZKP technology, and developer tooling and services.

Edited by Jonathan Stoker