Overview of Crypto & AI

Key Takeaways

• The Crypto x AI sector is growing rapidly, offering potential solutions to centralisation in the AI industry and providing an interoperable layer for AI agents to operate.
• Blockchain's role in AI includes increased access to global computation, transparent data exchange, and the potential for 24/7 permissionless AI markets.
• Challenges for Crypto x AI include scalability issues, data security concerns, and navigating the speculative market nature.

How Does Crypto & AI Intersect?

The intersection of crypto and AI creates a new landscape known as Crypto x AI. This collaboration leverages blockchain technology to provide solutions that ease the centralisation of AI resources, improve global access to computational power, and foster interoperability between different AI agents.

Blockchain can provide AI systems with secure, transparent data exchanges and a decentralised environment to operate. For example, decentralised marketplaces can enable AI agents to access computational resources on demand while maintaining a transparent record of transactions. Moreover, blockchains can help align incentives in AI projects by providing a trustless system for data and resource exchanges.

Why Does AI Need Crypto?

Crypto provides a secure, decentralised infrastructure for AI development, addressing key challenges faced by AI today:

• AI requires extensive computational power and data, which can be costly and centralised. Blockchain networks like Render facilitate global access to hardware and computation, allowing AI projects to tap into a distributed pool of resources.
• Crypto’s immutable ledgers ensure that data exchanges remain transparent and auditable, crucial for verifying the integrity of AI models.
• Decentralised AI agents can operate in permissionless markets, enabling round-the-clock access to services like trading (e.g. Uniswap) and computation (e.g. Render). This allows AI agents to operate autonomously and securely across various apps​.

Challenges For AI in Crypto

Historically, blockchains have not been known for their performance, especially when compared to centralised databases, which poses a problem for AI applications that require complex computations and handle large datasets. Efforts to scale blockchains are ongoing.

AI often requires access to sensitive data. While blockchains provide data integrity, ensuring confidentiality is a challenge. Crypto x AI projects will need to continually address these security concerns, balancing the need for private and public data, especially in early-stage development.

Finally, the speculative hype surrounding AI and crypto may attract dishonest actors who overstate the extent to which their projects leverage AI. This constantly raises questions about Crypto x AI projects, such as whether blockchain is actually needed for the project and if it truly makes the application better​.

Examples of Crypto x AI Projects

Bittensor (TAO)

Bittensor (TAO) is a blockchain that consists of multiple specialised networks (i.e. subnets) designed for specific use cases related to machine learning. For example, there are subnets for conversational AI, similar to ChatGPT, and others for AI-generated images. (Machine learning is a subset of AI focused on building systems that learn and improve from experience without being explicitly programmed.)

The main parties in the Bittensor network are miners and validators. Miners run the AI models posted to the Bittensor network, competing with one another to produce the best outputs in order to be compensated with TAO tokens. Validators review miners’ model outputs and verify their results. Over time, validators will reach an agreement on ranking these outputs.

Render (RNDR)

Render (RNDR) is a decentralised network that allows users to access computational resources for rendering and AI tasks. By utilising idle GPU capacity globally, Render aims to provide a cost-effective way to handle the intense computational demands that come with using and training AI.

Artists, developers, and AI researchers can pay in RNDR tokens to access the network's services. This decentralised approach helps enable a more equitable distribution of computational power and supports AI's growing demand for resources​.

As the graphic below from Galaxy Research illustrates, subse

Conclusion

The intersection of crypto and AI is a relatively new and fast-growing category, with dozens of projects developing solutions to current AI challenges while driving new opportunities for decentralised markets. As the Crypto x AI sector develops, projects like ASI and Render are leading the way, showcasing how blockchain technology can support AI's evolving needs. These AI crypto projects still need to demonstrate product-market fit and stand out from centralised AI projects as being better on the blockchain.