There has been a lot of buzz around AI agents and Web3. Many have concluded that AI agents are the future. My personal forecast is that we eventually see the end of the website. These are designed for humans.
AI does not need a website. In fact, it is likely they will be more efficient by simply gathering data from other agents. This is where the dead Internet theory starts to take on some validity.
The problem, so far, is AI agents tied to DeFi are slow to emerge. What is the hang up? We see the use of agent exploding in the enterprise world. Why are we not seeing it translate into Web3 platforms?
This is what we will undertake in this article.
Image generated using Ideogram
AI Agents And Web3: How To Create The Future
The reality of this entire discussion actually extends beyond the concept of DeFi. I think we can attribute it to Web3 in general. The progress outside "price go up" is rather slow.
Leaving that aside, the specifics pertaining to DeFi and AI agents come down to this:
With their infrastructure still in its nascent or development stage, the concept of DeFi-native agents will remain abstract for only so long. The AI economy now sits in a holding pattern—waiting for the build to catch up to the narrative.
If Web3 is excelling, it is in the financial arena. We do see a great deal of activity related to decentralized transactions. This led to the construction of many applications, which are expanding at a significant rate.
The issue, as we can see, is infrastructure. Underlying this is the appeal of AI agents and what they offer individuals along with businesses.
“While their potential in various application domains is still being explored, agent-based systems truly represent an immense opportunity for both companies and people”.
My view is this sentiment will only grow. The number of people seeing the potential is increasing. Individuals are expanding their usage of AI. As the technology is added to different platforms, people will simply start to utilize it. This is no different from the Internet. Early on, people tried to "learn" how to use it. Over time, the Internet itself became the infrastructure.
In other words, people to do use the Internet. Instead, they use applications or platforms which are built.
We are seeing another layer being laid over the top. By integrating AI into applications and platforms, people engaging with something else. At the root of this interaction is the AI agent, even if people do not realize it.
Eventually, this will be as natural for people as opening up an application. However, the process to get there, for DeFi, is still challenging.
The Future Path For Web3
There are a number of start ups looking to solve the AI agent problem relating to DeFi. The focus is upon infrastructure, seeking to offer up reliable data for the agents to utilize.
Unfortunately, this revealed another issue:
Interestingly, this also addresses a growing issue in DeFi: the fragmentation of AI agents. Handling token swaps, yield strategies, or cross-chain bridging often operates in isolation, with little to no coordination between them. The result is a disjointed user experience that’s difficult to navigate and scale.
We have an inherent problem with decentralization. Another term for that could be "fragmentation". For many, this is a positive, something I agree with. For example, it is hard to hack something when the server cluster is mirrored yet not directly connected under one corporate entity. What this means is accessing one server does not provide the ability to alter the others.
However, when it comes to things such as data feeds for oracles, this can be a problem. It is something the industry will have to work through.
AI agents survive on data. That is their food and a lot is required. If that can be solved, we could see agents such as the following being developed.
Here are some examples of AI agents for DeFi according to Rafiki.
- Trading Bots: AI-powered trading bots can analyze market trends and make predictions to optimize trading strategies.
- Risk Management Agents: AI agents can help identify and mitigate potential risks in DeFi investments, such as price volatility or smart contract vulnerabilities.
- Portfolio Optimizers: AI-powered portfolio optimizers can help users diversify their investments and maximize returns.
- Market Sentiment Analyzers: AI agents can analyze market sentiment and provide insights to inform investment decisions.
- Compliance Agents: AI-powered compliance agents can help ensure that DeFi platforms and users are meeting regulatory requirements.
This is helping with the automation, one of the promises of Web3. Due to the integration of blockchain, we could see a great deal of data opened up. Building infrastructure that accesses does not pose much of a challenge.
One obstacle that does arise is not all information related to DeFi is on blockchain. Certainly things such as actual token movement are available. But when looking at things such as regulations or sentiment, that requires information from outside blockchain. Here is where the present limitations of Web3 appear.
Therefore, once again, we revert back to the idea of having as much information on blockchain as possible. The transition from Web 2.0 has been slow. People still flock to the traditional platforms, providing massive amounts of data to Meta, X, and Google. These companies are doing a tremendous job building agents for their platforms. This does not, however, help DeFi or the decentralized future in any manner.
Nothing operates in a vacuum. It is all interconnected, especially when it comes to data. This is a point that should be clear to everyone.
The success of Web3 is going to depend upon the data. It is that simple. Without it, the ability to build is limited. Infrastructure can help with that, since it is the "pipe". However, as we expand beyond price feeds, other sources are required.