Blockchain as a Solution for AI Content Consent, Says Aptos Co-Founder Avery Ching

Artificial intelligence (AI) models training on user-generated content could benefit greatly from blockchain technology, according to Avery Ching, co-founder and CTO of Aptos. Speaking at the India Blockchain Week 2024, Ching emphasized that blockchain’s ability to provide transparent, immutable records makes it an ideal platform for managing consent around the use of content for AI training purposes.

Blockchain for AI Training Data Control

Ching discussed the growing intersection of AI and blockchain, noting that while AI models, especially large language models (LLMs), are receiving a lot of attention, he finds the underlying data used to train these models particularly intriguing. He highlighted a practical example: content created during events—such as video recordings or news articles—could be used to train AI systems. The question arises as to whether content creators should have control over whether their content is included in AI training datasets. This is where blockchain could play a pivotal role.

Ching explained that blockchain’s immutable nature and decentralized structure make it an ideal tool for managing consent in AI training. He suggested that, through blockchain, individuals could grant or deny permission for their content to be used for AI training, providing a clear, auditable mechanism for consent.

While blockchain offers an elegant solution, Ching also acknowledged the challenges that come with implementing this system at scale, particularly on platforms like Facebook, where billions of pieces of content are generated daily. Each piece of data would require a corresponding blockchain transaction to manage permissions, which could lead to significant scalability issues and high costs.

Ching noted that such systems would need high transaction throughput, which might be expensive to maintain. Despite these challenges, he expressed confidence in the potential of blockchain to support this kind of use case, particularly as AI and content creation continue to grow and evolve.

Aptos’ Role in AI Content Consent

Aptos, Ching’s blockchain platform, is actively exploring ways to integrate blockchain solutions for managing AI training data permissions. He sees blockchain as a “natural back end” for this kind of application and believes that, as AI continues to play a larger role in content creation and consumption, blockchain could become an essential tool for ensuring that content creators retain control over how their data is used.

This perspective aligns with growing concerns over data privacy, ownership, and the ethics of AI training, making blockchain an increasingly valuable technology in the ongoing discussion about AI’s impact on society.

In summary, Ching views blockchain not only as a way to streamline consent but as a fundamental technology to reshape how content is used in AI training. Despite scalability challenges, he sees its potential to become a key player in AI’s future development.