AI on Blockchain (Series)

Episode 1: Introduction

I recently spoke at Ai4 2023, the largest AI conference in North America, regarding how blockchain technology can provide real world value to AI development and deployment. The video of that speech will be released by Ai4 in the coming weeks, but in the meantime I wanted to get some thoughts down regarding how these two technologies are intersecting. This will be a series of posts listing the value proposition blockchain bring to AI, as well as some real world examples of how the Internet Computer can be, or will be, used for AI development and deployment. Be sure to subscribe to follow along with this series.

While it’s likely that other layer-1 chains will be able to host AI at some point, as far as I know the Internet Computer is the only blockchain capable of playing a role in AI as it stands today, so I will use “blockchain” and “Internet Computer” interchangeably throughout this series. It’s also important to state that this series is not meant to be a discussion of theoretical use cases between these two technologies, but rather specific use cases that are possible with the technology as it exists today.

Value Proposition

This series will focus on 4 major unique selling points that blockchain brings AI: security, digital ownership, decentralized control and transparency.

Security - The ability for people and companies to secure their data and model development without the fear of data leaking, privacy incursion or system hacks.

Digital Ownership - The ability for people and companies to cryptographically prove ownership over any AI component (data, model, code) and the ability to easily transact that ownership and define licensing agreements within the digital product itself.

Decentralization - The ability for people and companies to remove trust assumptions and risks associated with AI model development and deployment.

Transparency - The ability to build trust with AI model consumers by giving them insight, and possibly control, into the model inputs, processes and outputs.

These four value drivers for blockchain will comprise most of the use cases between blockchain and AI in my opinion.

Team Size

While web3 has been mostly led by solopreneurs and small developer teams, followed by later adoption by larger companies and organizations, I believe the reverse will happen for AI and Blockchain. During the Ai4 conference, I heard many AI companies speak on their pain points regarding the four unique selling points above. Security and transparency were particularly common concerns for Ai4 speakers. For that reason, I believe the revolution of AI to the blockchain will happen with existing AI service and software providers/companies moving some of their product’s infrastructure to a blockchain. It will be enterprises who first explore the space of AI on blockchain and who will first prove out the use cases and find product-market fit. From there I predict smaller developer teams will carve out niche markets in this space.

I have much, much more to say on this topic and things are evolving fast. Be sure to subscribe to remain in the loop.