I recently took a Solidity bootcamp course with Puerto Rico Blockchain Trade Association. For many non-techs, the crypto fad has died down in favor of AI. An interesting question that was posed was:

How can crypto and AI work together?

For anyone familiar with memecoins and NFTs, the obvious answer is generative assets. There have been a number of generative art NFTs using various algorithms as well as generative image tools like Midjourney or Stable Diffusion.

But looking beyond the low-hanging fruit, there are some really exciting use-cases here.

For the defi space, stable coins are the largest current use-case for crypto, include asset-backed as well as algorithmic stable coins. But we all remember the fiasco of Terra Luna, leaving a sour taste for algorithmic coins. But a novel use of AI would be in monitoring trading markets to help prevent the same thing from happening again. Even possible to identify early trends that weren’t planned for!

Besides securing algorithms for stable coins, AI could also help crypto in security by monitoring smart contracts and watching for unusual activity. Imagine if someone found a loophole in your smart contract in the middle of the night. Instead of relying on notifications and a manual contract upgrade to freeze it, you could have an AI agent monitoring for unusual activity, and even automatically freeze the contract. (Assuming your contract is upgradeable.)

You could also use crypto + AI for novel uses with oracles. Companies like OpenAI, Meta, and Google are struggling to monetize their AI offerings, and for many non-privileged people, paying $20 / month to access latest models may not be feasible, especially if you only need it for a one-off use. There’s also the question of capacity, with these AI offerings being hammered day-in and day-out by everyone and bots.

One implementation may be, locking your AI model behind a smart contract, where the user pays or redeems tokens to access it. These tokens could be earned via free-to-earn models such as surveys or ads; they could be time-gated and air-dropped; or purchased outright or traded on the market, allowing a secondary economy around AI. You could have “priority tokens” or NFTs that allow holders first access to the queue, or to early release models. The options are limitless!

But I think the big revelation is not so much how AI can help crypto, but how crypto and blockchains can help AI’s critical flaw: Safety and Attestation. There’s a very valid concern that generative AI assets can lead to deepfakes and disinformation, with no way to verify if they are true. While many people think of crypto strictly in terms of cryptocurrency, the blockchain is a powerful, trustless, decentralized tool for attestation and verification.

This isn’t just for AI generated content; there’s also a huge uncertainty with Digital Rights Management and copyright. You could publish proofs of authenticity for real videos or AI videos. One proposed solution was watermarking AI generated content, but this is hard to enforce. Perhaps ethical generative AI models will include a backend that automatically publishes the hash or an identifier to a blockchain (whether temporarily, using danksharding, or permanently).

There’s surely another 100 niche use cases for the intersection of crypto and AI, but I just wanted to share some of my insights.

This post is not officially endorsed or sponsored by PRBTA (yet!). You should check out their offerings and get involved.