AWS-Reliant Blockchains Won’t Bring Transparency to AI

The rapid progress of artificial intelligence (AI) has captivated the world, with many asking what is next for this technological breakthrough. While AI has already demonstrated its potential to transform various industries, it faces a major roadblock to large-scale adoption: a lack of trust and transparency.

Decentralized computation via the blockchain can alleviate the current trust issues, but there is a catch.

Dominic Williams is the founder and chief cientist of the DFINITY Foundation, a not-for-profit research and development organization and major contributor to the Internet Computer.

There is currently limited insight and no real way to validate the sources of data that an AI model is trained on, exactly what data is being collected by the model and, by extension, how that data informs the model and its accuracy.

Until there is a fundamental change to the transparency of AI programs and the infrastructure they are built on, users at all levels will not feel secure in leveraging these models due to a lack of trust and overall skepticism.

The intersection of AI and blockchain technology offers synergies that will enhance both technologies and drive widespread adoption through their integration.

See also: Inside the Orb: The Untold Story of Worldcoin's Launch

Currently, most blockchains lack the necessary infrastructure to support AI models due to their limited computing power as AI requires substantial computational resources and data sets. Limitations on computing power are caused in part by the fact that the majority of blockchains are not fully decentralized.

Instead, many of the world’s most popular blockchains today are reliant on a centralized cloud infrastructure (i.e. Google Cloud and Amazon Web Services), hindering the blockchain’s ability to support processing and storing data at the speed required for AI.

Despite negative headlines concerning the current attempts at integrating AI with blockchain are not what they seem. Current integrations have resulted in AI that runs WITH the blockchain rather than the desired goal of running AI ON the blockchain.

The core infrastructure and underlying technology for these “blockchain AI” projects mainly operate on centralized servers, and utilize plugins that connect centralized AI models to blockchains running on centralized cloud networks. This defeats the purpose of leveraging blockchain technology for AI as it doesn’t address the underlying issues of trust and transparency.

A completely/fully decentralized blockchain, such as The Internet Computer (ICP), the network I helped build that offers compute power matching or exceeding Web2 cloud servers, will allow AI models to be run entirely inside of smart contracts. This will make the training parameters and inputs that make large language models both open source and tamper-proof. To enable AI integration on the blockchain, we need blockchains capable of processing data at speeds comparable to Web2 clouds, which can only come from full decentralization.

Hosting AI models on the blockchain itself allows AI systems to leverage inherent decentralization to increase the transparency of every aspect of the model. Thus, AI on the blockchain is the next logical step for long-term success as blockchain will enhance AI’s credibility, accountability and security, fostering greater trust among users.

However, there are misconceptions about how exactly the two pieces of technology can operate in unison and until they are dispelled, the growth of the AI ecosystem will be unable to reach its full potential.

Fully realizing the potential of AI on the blockchain requires a truly decentralized network. It must be capable of storing and processing data so that full models can be run unencumbered within smart contracts. These decentralized systems like ICP will empower AI to function as an autonomous cloud, transforming the landscape of AI development.

For example, consider an AI model designed for medical professionals. The model is widely used but ultimately produces untrustworthy responses. That is because there is no easy way to verify the training data the model was built on and how that data was used.

This centralized model only produces outputs, with no insight into the inputs. However, in a decentralized environment, the AI large language model can be built solely off of renowned medical textbooks and reputable databases of medical research papers.

When the doctor interacts with the AI, the hidden process is fully transparent and the cryptographic proof guarantees what content the AI has been trained on. Thus, the generated response can be verified and doctors can trust the results.

This example is just one of many that demonstrate why decentralization is critical to building trust in AI models. By operating in a fully open and public environment, AI on the blockchain ensures transparency in data processing, enabling users to understand how their data is utilized.

Further, on-chain AI applications can all access and contribute to the same data set, creating a collaborative ecosystem within the blockchain. The tamper-proof and secure nature of blockchains ensures that this data is also less susceptible to misuse for malicious purposes.

See also: 10 Ways Crypto and AI Can Make Each Other Better

The collaboration between AI and blockchain presents a remarkable opportunity to advance both technologies and promote a more trustworthy and reliable exchange of information.

By addressing concerns around trust and transparency this integration offers a more reliable and transparent digital ecosystem. The potential of AI on the blockchain is vast, promising seamless integration, open-source tamper-proof smart contracts and rapid content generation for metaverses, games and decentralized social media.

The future holds immense possibilities at the intersection of AI and blockchain, driving us toward a decentralized future.