
In just over a year since the debut of ChatGPT, generative AI has become a significant global phenomenon. OpenAI’s early success ignited a surge of investor interest in large language models (LLMs) and AI applications, attracting a whopping $25 billion in funding in 2023 alone—a fivefold increase from the previous year! This enthusiasm underscores the enormous potential of this multi-trillion-dollar market.
AI and cryptocurrency technologies naturally complement each other, fostering a growing ecosystem of AI within the web3 space. Despite the buzz, there is still a lot of confusion about the roles of various protocols, what’s hype versus reality, and how these elements fit together. This article aims to demystify the web3 AI supply chain, break down the tech stack, and explore the competitive landscape. By the end, you’ll have a clear understanding of how the ecosystem works and what to look out for next.
I. Web3 AI Infrastructure
Generative AI Workloads: Training, Fine-Tuning, Inference
Generative AI relies on LLMs, which run on high-performance GPUs. These LLMs handle three main workloads: training (model creation), fine-tuning (specializing models for specific sectors or topics), and inference (executing the model). The infrastructure layer is segmented into general-purpose GPU marketplaces, ML-specific GPU marketplaces, and GPU aggregators, each catering to different workload capabilities and use cases.
General-Purpose GPU Marketplaces
General-purpose GPU marketplaces, such as Akash and Render, provide decentralized, crypto-incentivized computing power suitable for any application, with a focus on model inference—the most common LLM workload. These marketplaces are positioned to grow exponentially as AI becomes more integrated into daily life. Although computing resources are technically a commodity, demand for decentralized, GPU-specific compute in web3 is expected to rise. Long-term success will hinge on distribution and network effects.
ML-Specific GPU Marketplaces
ML-specific GPU marketplaces cater to machine learning applications, supporting model training, fine-tuning, and inference. These platforms can differentiate themselves through ML-specific software overlays, but like general-purpose marketplaces, distribution and network effects are crucial. Bittensor is a frontrunner in this category, with many new projects on the horizon, signaling significant market potential.
GPU Aggregators
GPU aggregators, such as Io.net, combine GPU supply from general-purpose and ML-specific marketplaces, simplifying networking orchestration and integrating ML-specific software. These platforms offer comprehensive GPU solutions capable of handling model training, fine-tuning, and inference. As the demand for consolidated GPU distribution grows, more competitors are expected to emerge.
II. Middleware Solutions
Bridging GPUs and Blockchain
While the infrastructure layer provides access to GPUs, middleware is necessary to connect these computing resources to the blockchain in a trust-minimized manner. This is where zero-knowledge proofs (ZKPs) come into play.
Zero-Knowledge Proofs (ZKPs) for AI
ZKPs are cryptographic methods allowing one party (the prover) to prove to another party (the verifier) that a statement is true without revealing any additional information. For AI, the “statement” is the LLM’s output given specific input. Decentralized marketplaces for ZKP verifiers bid on verifying inference outputs, ensuring accuracy while maintaining data and model privacy. Though zkML (ZKPs for machine learning) is in its early stages, advancements in this area could revolutionize web3 and AI, opening new design spaces and use cases. Leading developers include =nil=, Giza, and RISC Zero, with protocols like Blockless acting as aggregation and abstraction layers for ZKP distribution.
Developer Tools and Application Hubs
In addition to ZKPs, developers need tools, SDKs, and services to build AI applications such as AI agents and automated trading strategies. These protocols often double as application hubs, providing platforms for users to access finished applications. Bittensor and Fetch.ai are notable leaders, offering comprehensive platforms for developing and distributing enterprise-grade AI solutions.
III. Application Layer
Real-World Use Cases
At the pinnacle of the tech stack, user-facing applications leverage web3’s permissionless AI processing power to perform specific tasks across various industries. Although still in its nascent stages, the potential applications are vast and growing. Here are some notable examples:
Smart Contract Auditing
Smart contract auditing involves analyzing and verifying the code of smart contracts to ensure their security and correctness. AI can automate and enhance this process by identifying vulnerabilities and suggesting optimizations, significantly reducing the risk of hacks and errors. With web3 infrastructure, these audits can be conducted in a decentralized manner, ensuring transparency and trustworthiness. This can be particularly beneficial in financial applications where the integrity of smart contracts is paramount.
Blockchain Chatbots
Blockchain chatbots provide automated, intelligent interactions within blockchain environments, enhancing user experience and engagement. These chatbots can facilitate customer support, guide users through complex processes, and even execute transactions based on conversational inputs. By leveraging AI, these chatbots can understand and respond to natural language queries, making blockchain technology more accessible and user-friendly. Additionally, they can operate securely within the decentralized framework of web3, preserving user privacy and data integrity.
Metaverse Gaming
The metaverse represents a burgeoning field where AI and web3 intersect to create immersive, interactive virtual worlds. AI can drive dynamic content generation, character behavior, and adaptive storylines, making gaming experiences more engaging and personalized. Web3's decentralized nature ensures that these virtual worlds are not controlled by any single entity, promoting creativity and innovation. Gamers can own and trade in-game assets securely using blockchain technology, creating real-world value and fostering vibrant virtual economies.
AI in Trading and Risk Management
In the financial sector, AI is transforming trading and risk management by analyzing vast amounts of data to identify patterns, predict market movements, and make informed decisions. AI algorithms can execute trades at high speeds and optimize investment strategies, enhancing profitability and reducing human error. Within web3, these AI-driven trading systems can operate transparently and autonomously, leveraging blockchain's immutable record-keeping to ensure accountability and compliance. Additionally, decentralized finance (DeFi) platforms can utilize AI for risk assessment, fraud detection, and regulatory compliance, improving the overall security and efficiency of financial markets.
Future Prospects
As the underlying infrastructure and middleware technologies advance, we can expect to see the emergence of next-generation AI applications with functionalities that are difficult to imagine today. Here are some anticipated innovations:
Enhanced Privacy and Security
Future AI applications will likely incorporate advanced cryptographic techniques, such as homomorphic encryption and differential privacy, to enhance user privacy and data security. This will enable AI to process sensitive information without exposing it, fostering trust and broadening AI adoption across industries like healthcare, finance, and personal data management.
Decentralized Autonomous Organizations (DAOs)
AI can play a crucial role in the evolution of DAOs, which are member-owned communities governed by smart contracts. AI algorithms can help DAOs make more informed and efficient decisions by analyzing data, predicting outcomes, and automating routine tasks. This can lead to more resilient and adaptive organizational structures, capable of responding to changing circumstances in real-time.
AI-Driven Content Creation
The intersection of AI and blockchain can revolutionize content creation across media, art, and entertainment. AI can generate high-quality text, images, music, and videos, while blockchain can ensure the authenticity, ownership, and monetization of digital content. This synergy can empower creators, reduce production costs, and open up new avenues for artistic expression and revenue generation.
Autonomous Agents and IoT
AI-powered autonomous agents can interact with the Internet of Things (IoT) devices to perform tasks autonomously, ranging from home automation to industrial processes. These agents can leverage blockchain to ensure secure communication and coordination between devices, creating more efficient and resilient systems. For instance, in smart cities, AI agents can manage energy consumption, traffic flow, and emergency response, enhancing urban living conditions.
Market Dynamics: Early Entrants vs. New Leaders
The market for web3 AI applications is highly dynamic, with new entrants constantly emerging. While early players have the advantage of first-mover status and established user bases, they must continuously innovate to maintain their lead. New entrants, on the other hand, can capitalize on the latest technological advancements and learn from the experiences of early adopters to introduce disruptive innovations.
Sustaining Innovation
For early entrants to sustain their market position, they need to invest in research and development, expand their feature sets, and adapt to evolving user needs. This might involve integrating newer AI models, enhancing user interfaces, or expanding into new use cases.
Disruptive Innovation
New players can disrupt the market by introducing groundbreaking technologies or novel business models. They might focus on niche markets or specific pain points overlooked by incumbents, leveraging their agility to quickly adapt and scale. Collaborations with established tech firms and strategic partnerships can also accelerate their growth.
In conclusion, the application layer of web3 AI holds immense potential to revolutionize various industries by providing innovative, decentralized solutions. As technology advances, we can expect to see even more sophisticated applications that push the boundaries of what is possible, creating new opportunities and transforming the way we interact with digital systems.
IV. Investor Perspective
The rapid development of AI and web3 technologies presents a compelling landscape for investors. Understanding where to allocate resources within this burgeoning field can significantly impact returns. This section explores investment opportunities within the AI tech stack, emphasizing the infrastructure and middleware layers, and provides long-term market predictions.
1/ Investment Opportunities in AI Tech Stack
Infrastructure Layer: The Backbone of AI
The infrastructure layer is critical as it provides the computational power necessary for AI operations. This includes general-purpose and ML-specific GPU marketplaces, as well as GPU aggregators.
- General-Purpose GPU Marketplaces: Investing in platforms like Akash and Render can be advantageous as they cater to a broad range of applications. The growing demand for decentralized computing power ensures a steady market for these services. Investors should look for platforms with strong distribution networks and user bases.
- ML-Specific GPU Marketplaces: Platforms like Bittensor, which cater specifically to machine learning workloads, offer differentiated value by integrating ML-specific software. These marketplaces are poised for growth as more businesses seek specialized AI solutions. Investors should focus on projects that demonstrate technical superiority and robust ML integration.
- GPU Aggregators: Companies like Io.net provide comprehensive solutions by combining general-purpose and ML-specific GPU resources. These aggregators simplify the user experience and can capture a larger market share by offering end-to-end solutions. Investment in these platforms could yield high returns as demand for streamlined AI infrastructure grows.
Middleware Layer: Enabling Seamless Integration
The middleware layer connects computational resources with blockchain, ensuring secure and efficient operations. This layer includes zero-knowledge proofs (ZKPs) and developer tooling and application hubs.
- Zero-Knowledge Proofs (ZKPs): Investing in ZKP providers like =nil=, Giza, and RISC Zero can be lucrative, as these technologies are essential for verifying AI outputs while preserving data privacy. As zkML (ZKPs for machine learning) evolves, these companies are likely to see increased demand. Platforms like Blockless, which act as aggregation and abstraction layers, offer additional investment opportunities by providing versatile solutions across various ZKP providers.
- Developer Tools and Application Hubs: Platforms like Bittensor and Fetch.ai are critical for developers building AI applications. These hubs offer the necessary tools and SDKs, making them attractive investment targets. Their dual role as application distribution platforms enhances their value proposition, providing multiple revenue streams and broad market appeal.
2/Long-Term Market Predictions
Growth Trajectory and Market Dynamics
The AI and web3 market is expected to grow exponentially, driven by technological advancements and increasing adoption across industries. Here are some key predictions:
- Exponential Growth in Demand: As AI becomes more integrated into daily life and business operations, the demand for AI infrastructure and middleware will skyrocket. Investors can anticipate significant returns by backing companies that provide the foundational technology for these applications.
- Emergence of Dominant Players: While the market is currently fragmented, dominant players are likely to emerge, particularly in the infrastructure and middleware layers. These companies will benefit from economies of scale, network effects, and robust distribution channels. Investors should identify and support platforms with the potential to achieve market leadership.
- Innovation in AI Applications: The application layer will see continuous innovation, with new use cases and functionalities emerging. While investing in application startups can be riskier due to the uncertainty of market adoption, successful applications can offer substantial returns. Monitoring trends and backing innovative projects with strong technical teams and viable business models will be crucial.
Focus on Infrastructure and Middleware
Given the foundational nature of infrastructure and middleware, these layers present more stable and potentially lucrative investment opportunities. Here’s why:
- Stability and Predictability: Infrastructure and middleware services are essential for AI operations, ensuring a steady demand. These layers are less susceptible to market fluctuations compared to the application layer, where user preferences can shift rapidly.
- Scalability and Integration: Infrastructure and middleware platforms can scale efficiently and integrate with various applications, broadening their market reach. Investments in these layers can benefit from the overall growth of the AI ecosystem, regardless of specific application trends.
Long-Term Market Predictions
- Infrastructure Expansion: As the need for high-performance computing grows, infrastructure providers will expand their capabilities, leading to increased investment in data centers, advanced GPUs, and other hardware. This expansion will support more complex and demanding AI workloads, driving further innovation.
- Middleware Advancements: The middleware layer will continue to evolve, with improvements in ZKP technology and developer tools making AI integration more seamless and secure. These advancements will open up new possibilities for decentralized AI applications, enhancing their functionality and appeal.
- Regulatory Considerations: As AI and web3 technologies mature, regulatory frameworks will become more defined. Investors should stay informed about regulatory developments, as compliance will be crucial for the long-term success of these technologies. Companies that proactively address regulatory challenges and build compliant solutions will be better positioned for sustainable growth.
- Global Market Expansion: The adoption of AI and web3 technologies will expand globally, with significant growth opportunities in emerging markets. Investors should look for companies with strategies to penetrate these markets, leveraging local partnerships and adapting to regional needs.
Conclusion
Investing in the AI tech stack, particularly in the infrastructure and middleware layers, offers substantial opportunities for growth and returns. By focusing on foundational technologies that enable a wide range of applications, investors can position themselves to benefit from the exponential growth of AI and web3. Monitoring market trends, supporting innovative projects, and staying informed about regulatory developments will be key to making informed investment decisions in this dynamic and rapidly evolving field.