Over the past year, we have seen rapid growth in the capabilities and applications of artificial intelligence in all areas, including the crypto industry. The integration of artificial intelligence (AI) and blockchain has become one of the leading topics, as every day there are new applications of this synergy in the industry and how they can improve each other.
In a recent publication, the cryptocurrency exchange Coinbase highlighted the latest breakthroughs in artificial intelligence and their intersection with blockchain technology. The current article reveals the benefits of combining blockchain and AI and the main points of the Coinbase publication.
Coinbase relies on Bloomberg’s forecast, noting that by 2032, the generative AI market alone will be worth $1.3 trillion, with a CAGR of 42% over the next decade.
The publication highlights two main aspects of the intersection of AI and blockchain:
Using AI/ML models and techniques to improve blockchain platforms, decentralized applications, and trading digital assets on the blockchain.
Leveraging the power of blockchain to create value for developers and users of AI and ML products and services. Coinbase emphasizes that this aspect is the main topic of the document.
They point to a few fundamental blockchain capabilities that can be used to develop, deploy, and operate an AI model. The capabilities provide the following key properties that are important for the AI use cases described in the paper. For example:
Data security: Helps to store data in a secure and unaltered form, with high availability, using decentralized servers that are less susceptible to attack
Data provenance, traceability, auditability: Recording transactions and assets in an immutable and transparent way that facilitates tracing the source, ownership, and origin of data and transactions through digital signatures and time stamps. It also enables control and verification.
Decentralized decision-making: The ability to make decisions between multiple entities in a context where they do not have established agreements between them or with a central authority.
Autonomous and transparent code execution: Ensuring the execution of programs in the form of smart contracts that are transparent to all stakeholders and work autonomously, without the need for centralized intermediaries.
Decentralized identities: Creating secure digital identification mechanisms that allow users to interact with services without compromising their privacy.
Micropayments: a secure and easy way to make payments at low cost.
Coinbase has divided the potential benefits of blockchain for AI and related products into four main categories:
Experts note that blockchain can be used to create solutions that will help users and developers ensure that data and models have not been altered without their knowledge.
As an example, they mention an API-based service that can allow data owners and AI developers to record time-stamped hashes of datasets and models to ensure their integrity, as well as record the entire process of developing models and datasets.
The system can even be directly integrated into ML development tools such as Pytorch. Such integration would help improve the integrity and reliability of models, making the model development process more transparent and secure.
Furthermore, Coinbase notes that registering data hashes and modeling results on the blockchain can also help prevent deep fakes. For example, applications can ensure the authenticity of the data used by verifying digital signatures associated with the data source on the blockchain, or a decentralized version of the Snopes.com website can be developed and implemented on the blockchain to flag deepfakes.
Analysts emphasize the role of NFTs, noting that the token can demonstrate ownership of any digital content or data. Depending on the use case, the content can be the input to a model, for example, a hint for a generative artificial intelligence tool. The NFT will allow a user or developer to claim ownership and subsequently transfer ownership of the relevant digital asset to another person.
Coinbase believes that it is also possible to implement a blockchain-based access control mechanism for data and models, such as a smart contract that allows/restricts access based on a predefined list of user addresses. Alternatively, it can be integrated with a decentralized identity solution to allow access based on certain provided data while maintaining user privacy.
Experts believe that the blockchain will be able to facilitate micropayments for the use of a generative AI model using low-fee stablecoins, and the smart contract may allow for decentralized revenue sharing among several model co-owners. Such co-ownership models can allow small and medium-sized model developers to join forces and compete with large firms in the industry.
Additionally, blockchain can be used to create a decentralized data marketplace that makes it easier for model developers, users, and others to find and match with each other, offer rewards, make payments, and enter into contracts and agreements. A decentralized blockchain-based verification system implemented through smart contracts can incorporate both automated and human data verifiers into a system that encourages high throughput, thorough and high-quality data, and model verification.
This includes running certain AI models directly on the blockchain to benefit from greater transparency and trust. AI models will be able to perform certain data processing functions for end users, giving them the confidence that only the model they assigned the input to received it and generated the result they see, without any manipulation, tampering, or censorship. Or they can be used to help the smart contract adapt and optimize its parameters in response to user transactions.
Moreover, the models can be owned by individuals or decentralized in the form of DAOs, which allows multiple individuals and legal entities to own “stakes” in a particular on-chain model. In the long term, for such applications, given the high data and computing requirements, it may be interesting to research and develop a new platform designed from the ground up to support AI workflows.
In conclusion, Coinbase shares its view on fostering this dynamic development by striving to increase economic freedom around the world by building reliable, interoperable products and supporting developers in the generative AI ecosystem.
The company has expressed interest in partnering with organizations that have experience with AI solutions to bring these innovations to life.
Coinbase is not the only company that is open to the introduction of artificial intelligence into its work. For example, WhiteBIT CEO Volodymyr Nosov said in a recent interview that the company has also implemented some AI solutions.
“Analytics in our company is based on a large amount of data, so we use specialized tools. They are based on well-known artificial intelligence models but adapted to our needs. In addition to the obvious, they also help us develop products in our ecosystem. We see their effectiveness, so we are always ready for new integrations,” Nosov said.
We can also mention the recent successful experience of Vitalik Buterin. The expert explained how the successful integration of AI into Ethereum helped to fix technical errors in the blockchain code base. Buterin has repeatedly expressed optimism about this combination.
“Now that both blockchains and AIs are becoming more powerful, there is a growing number of use cases in the intersection of the two areas. However, some of these use cases make much more sense and are much more robust than others.” Buterin wrote, adding in the end “I look forward to seeing more attempts at constructive use cases of AI in all of these areas, so we can see which of them are truly viable at scale.”
Coinbase’s publication emphasizes the potential benefits of using blockchain for AI and its products. Moreover, as we can see, numerous companies have already started to implement artificial intelligence in their operations. In general, the integration of blockchain and AI can bring significant benefits, including increased data integrity, incentives and payments for the use of data and models, as well as greater transparency and trust in the work of artificial intelligence.
Deniz Tutku