Skip to content

Web3 Magic: How AI Agents and Blockchain Reshape Tech

AI Agents

Artificial Intelligence (AI) and Blockchain technology. Just a few years ago, these were just “buzzwords” with little to no real world application. Things have changed drastically since then.

In the wake of ChatGPT’s transformative impact on bringing AI into the mainstream, the once abstract realm of AI and Machine Learning has now become widely recognized for its inherent value. 

The overarching question has shifted from “What can AI do” to “What can AI NOT do to safeguard our jobs” Safe to say, almost everyone believes that AI is going to take over our lives. For something so life-changing, it is important to handle this technology properly. 

However, there is a pressing concern surrounding centralization of this technology. We can see this in the power struggle between Sam Altman, the CEO of OpenAI, and the board. Till today, we are still unclear about what happened and what led to the power struggle.

Do we really want to leave our fates to the select few elites? 

Fortunately, there is a better way. By using Blockchain technology to govern AI agents, we can ensure that AI agents are managed by multiple parties. 

In this article, we will explore the utility of AI agents, why they will change our future, and why we need to use Blockchain technology to manage AI agent solutions. 

What are AI agents and why do they matter? 

AI agents operate with minimal human input, making them instrumental in various applications. Think of it as hiring an assistant to do all your tasks for you. Now instead of a human being, AI agents are softwares that you can train and use to execute your tasks for you.

AI agents are birthed through the consumption of vast datasets by Large Language Models (LLMs), allowing them to autonomously perform specific tasks.This best example of this: ChatGPT.

Photo by Matheus Bertelli: 

ChatGPT is an AI powered LLM that consumes huge volumes of data scraped from the internet and content that it was fed. The LLM algorithm interprets the data and builds connections between the different types of content. In other words, ChatGPT understands the context of your question using the data sets it consumed and gives you an answer. 

However, AI agents are not just limited to “answering machines” that reply to your queries. They can be used in applications such as accountancy and book-keeping, sorting voice transcripts, and even navigating the complexities of driving cars, just to name a few. 

As the technology improves and uthe ser experience gets better, using AI agents will be as common as using a cooking pan to fry an egg or driving a car to get from one place to another. They are all tools we use in our daily life. However, there is an issue that threatens to make AI agents dangerous.

Centralization Threatens To Derail The Future Of AI Agents

Imagine this for a moment. There is a button that, once pressed, will mean the end of the world. Will you want that button to be controlled by one good natured man? Or will you prefer if the button is managed by a group of people from different countries that bring together different view points in discussions?

I think you would prefer the latter.

Such an important technology such as AI must not be controlled by a select few companies or individuals. Here’s why blockchain technology must be integrated with Autonomous services for decentralized governance and more.  

Bringing everything on-chain using Blockchain

Firstly, the blockchain’s foundational principle of everything being on-chain and verifiable brings unparalleled transparency and trust to the ecosystem. Any changes made to the AI agents must be voted on, with transparent results that are easily verified. This facilitates the governance of AI technology by the community. 

No one entity has full control of all the AI agents. The best ideas can be shared amongst the community and submitted as proposals for everyone to vote on. 

The performance of AI agents can be verified on-chain, with the best-performing AI agents compensated accordingly. This incentivizes creators to build better AI agents, which attracts more users, leading to better compensation for the creators. Creating a flywheel effect. 

Besides the governance of AI agents, Blockchain technology can be used to verify the quality of AI models.

Using zero-knowledge proof to verify model quality

Many AI agents may be run on models of low quality and integrity. For example, the data used may not be “clean” or the algorithm is not suitable for the problem. Retail users are not able to assess an AI agent’s suitability (unless they have a degree in data science and have access to the model) 

However, requiring AI models to be openly verified by the community is unfair to the creators as well. After spending time and effort building the models, it can be easily copied if everyone has access to it. We need a solution that can verify the AI agent’s suitability for the task without needing to share the code for the community to check. 

Simply put, AI models like ChatGPT is a “Black Box” for most of us. We only know what we input (our queries) and the output (the replies we get back) What happens in between is a big question mark, which is worrying as we do not know when this “Black Box” is manipulated.

Using Zero-Knowledge Machine Learning (zkML) technology is a potential solution to this problem. Zero-knowledge proof is a cryptography method that allows the prover to prove to the verifier that something is true without outright showing the evidence. 

For example, Alice wants to borrow $100,000 from Bob. But Bob will only agree to lend the money if Alice can prove that she has $100,000 in her Crypto wallet. Zero-knowledge proofs allows Alice to show that she has access to a wallet with at least $100,000 without sharing sharing her private key with Bob.

zkML technology allows the user of the AI agent to verify that a specific input led to a specific output using a particular model without revealing what the “Black Box” in between is. The verification can be done without revealing the AI agent’s model.

This ensures that you are using the correct model to get your desired outputs at all times. 

How To Assess Web3 Projects In The Autonomous Services Space

Are you looking to invest in Web3 projects that work in the AI realm? Unfortunately, lots of projects only have a small AI product feature but claim to be an “AI solution”. They do this to jump on the “AI hype train” to get funding and users in the near term.

Examining successful Web3 projects in AI agents, a common thread emerges. The most promising projects usually have these points in common

  • They operate with everything on-chain and prioritize decentralization. All actions completed by the AI agent are shown on-chain and the project is governed through on-chain voting 
  • High level of commitment to data quality
  • Able to attract a community of developers and data scientists that actively contribute to the project
  • Can explain key concepts of their technology to get retail investors excited

Every Project Wants To Attract Developers, But Few Do It Well

If you have the best developers building your product, chances are your product will turn out great. With the best product, users are more likely to give raving reviews and promote it to their contacts. 

No wonder every Web3 project is fighting for the best developers! But few actually succeed. 

Besides generous compensation, developers are looking for a promising story. They want a reason to be excited about building your project. Here’s what you can do.

Define Your Niche

Firstly, the key to success lies in a strategic and targeted approach to attract developers. Rather than trying to be everything for everyone, focus on providing a niche solution to be different. Specialization not only makes your platform more attractive to potential developers, but it also simplifies the vision, avoiding confusion. 

For example, machine learning technology can be useful for different industries, from shipping to real estate. Positioning your product as an “all-in-one” solution that can be used in 3 different industries will only confuse potential developers. Instead, focus on solving one problem for one industry. 

Show, Don’t Tell

Secondly, showcase and demonstrate your technology effectively – provide a hands-on experience that goes beyond mere theory. Offering a “free taste” allows developers to familiarize themselves with the capabilities of your project, fostering trust and engagement. While the best developers should easily understand the technology, giving them something they can “touch and feel” will help them “get” your project really quickly. 

Build Word of Mouth

Lastly, double down On “Earned” Media. Many people in the Crypto space are skeptical of paid ads or paid advertorials. Unfortunately, a big reason for this was the many scams in the past where investors got rug-pulled. Word of mouth and referral go a long way in making your project look credible in the eyes of investors and users.

To achieve “Earned” media, you only need two things – 1) having the right story and 2) reaching out to the right Key Opinion Leaders (KOLs)

Sounds simple, but few get it right.

The “right story” needs to be different and simple to understand. Different in terms of the product and the vision of the project. Without a different story, you will be seen as “just another project” and builders will prefer building for the leader instead. 

However, do not get lost in sharing how amazing your tech is when trying to be different. Simplify your product such that someone who has never heard of Crypto can understand it. 

After you have gotten the right story, start reaching out to KOLs. More importantly, reach out to the right KOLs. For example, if you are a staking project, reach out to KOLs we consistently write about their staking experience or how they maximize their yields. Do not reach out to NFT influencers, even if they have a strong following. Sounds obvious, but many projects still make this mistake.

Share with these KOLs about your project, why you are different and offer them exclusive early testing and ask for their feedback. Try not to offer incentives early on and get their real feedback first. You want them to be evangelizing your product because they truly believe in it, rather than because they are paid to do so. If they are not excited about your project without the incentive, that is feedback for your product. 

Build Up Hype For Your Project Using This Playbook

Are you a founder building an AI and/or Blockchain product? Do you struggle to explain how awesome your product is? Or is it hard to get the Web3 community talking about your project (at least in a positive way)?

I can help you create and execute a content playbook. Use this playbook to get your project noticed and attract users at scale.
Submit the contact form or email me at

Leave a Reply

Your email address will not be published. Required fields are marked *