Original author: SAURABH DESHPANDE Compiled by: LlamaC

What if everyone could have a private banker with just one button? What if the banker could hire a large army of analysts, compliance and executives to trade for you? It sounds a bit out of reach, but that's exactly what Saurabh explores in today's article. We are heading towards a world where robots move more money than humans. If Trump continues to serve as president, we will continue to see more assets being tokenized. Saurabh’s story today explores how the proxy economy in cryptocurrencies collides with the future of finance.

If you told someone in 1995 that decades later, they would be able to order meals, call a taxi, or transfer money to friends around the world, they might be skeptical. However, we are in an era where smartphones have reduced these once-complex tasks to simple clicks on the screen.

DeFi is now at a similar turning point. DeFi offers opportunities to earn earnings and discover new tokens early, but it is too complicated to use for most people. Managing your wallet, navigating across different blockchain networks, and understanding smart contract interactions feels like learning a new language. Furthermore, many are hesitant about participating in DeFi due to regulatory uncertainty. It is not surprising that DeFi accounts for only 10-20% of spot trading volume on centralized exchanges (CEXs). This is because CEX is easier to use and has clearer supervision.

This article explores how artificial intelligence can transform DeFi from a complex ecosystem serving thousands of people to an accessible financial platform serving millions of people. We will examine how AI-driven interfaces can gradually begin to bridge the gap between the huge DeFi opportunity and the average user’s need for simplicity. While all DeFAI (DeFi and Artificial Intelligence) applications are in their infancy, they show what DeFi can be: providing a smooth experience when interacting with financial instruments, from automated trading strategies to conversational interfaces that make complex trading feel natural.

Let's start with how financial markets are integrated with computers and algorithms for the first time. Since the 1980s, algorithms have begun to become part of the financial market in a meaningful way. They are the cornerstone of the modern market. From stock trading to currency exchange.

Algorithm and goldFinancial Market

Jim Simmons came to my mind when I thought of algorithms in the financial environment. The word "Legend" is easily placed before his name. He founded Renaissance Technologies, a U.S. investment company that changed the rules of the quantitative trading game. Its flagship fund Medallion achieved an eye-catching CAGR of 39% over 30 years (1988-2018).

To understand how extraordinary this is: $100 invested in the Medallion fund will grow to $2.1 million in 30 years, compared to $1,014 investing in the S&P 500. This difference is almost incomprehensible.

But the real magic is how they do it. The team at Renaissance Technologies is not working with Wall Street veterans, but is composed of PhDs in mathematics, physics and other hard sciences. Their approach relies entirely on mathematical models and algorithms to trade markets—a testament to the power of data-driven decision-making.

This focus on algorithms is not limited to hedge funds. In traditional financial markets, trading is becoming increasingly dependent on algorithms. A recent article pointed out that more than 75% of daily foreign exchange spot trading, or $5.6 trillion, are now done through algorithms. These systems reshape the trading desk, shifting the focus from human intuition to automated decision-making.

DeFi is still in its infancy in terms of automation. In contrast, algorithmic trading has existed in the traditional financial field for more than 30 years. Since 2020, the same data-driven revolution that changed Wall Street has also begun to knock on the door to DeFi.

Algorithms and DeFi

Decentralized exchanges (DEXs) and lending protocols became the fundamental pillars of this new financial ecosystem in 2020.

DeFi really became active when Compound started its liquidity mining program, which triggered an outbreak of activity. Aave’s TVL and prices soared around the same time. Several new revenue farms are launched every day. These farms offer great benefits, usuallyPay in native tokens of the protocol. However, the value of these gains is directly pegged to the token’s market price, adding a layer of complexity to the returns. I remember Sam Bankman-Fried said in an interview—

Imagine a magical box that does nothing, but people are investing millions in it because...why not? As more and more money piles up, the box becomes valuable—because everyone agrees that it is valuable. At some point, experienced traders come in and say, wow, look at all the money in this box! It must be a great box! The cycle continues like this—until, of course, it doesn't continue anymore.

This dynamic creates a differentiation. Savvy traders thrive, wandering between farms, profiting from tokens, and taking advantage of opportunities. Meanwhile, inexperienced participants struggle to understand the importance of continuing profit in such a volatile market. It is obvious that this version of DeFi is not designed to extend beyond a niche audience.

As the ecosystem expands, the need for tools to simplify DeFi interactions has become increasingly urgent. Lending and lending agreements continue to increase, creating demand for aggregators.

Yearn Finance was launched in February 2020, with a total lock-in volume of 2.5 million ETH (about $7 billion at the time). This is a turning point in DeFi's development.

It introduces an automated vault to optimize on-chain benefits and provide users with a clear risk-reward overview. These vaults allow users to deposit assets—stable coins, ETH and specific tokens—while DeFi experts propose and implement earnings strategies. Funds are then deployed into the entire DeFi ecosystem based on these strategies, with profits shared among users, platforms and strategy creators (especially acting as fund managers).

This model is an advancement in DeFi. DeFi feels accessible for the broader audience for the first time. Yearn removed most of the manual work needed to participate in the ecosystem while adjusting incentives among stakeholders. Here is a glimpse into what DeFi’s next iteration might be: efficient, user-friendly, and scalable.

While Yearn makes DeFi easier to use, its limitations become apparent as the ecosystem develops. On-chain earnings are beginning to normalize, Yearn’s strategy struggles to maintain its advantage. Key innovators such as Andre Cronje and the severe market conditions in 2022 caused TVL to plummet from its peak to approximately $250 million.

Yearn is the first project in the DeFi field to primarily attempt to automate earnings optimization, improving manual earnings farming by allowing users to delegate funds to experienced managers. But it still relies on human decision-making. Strategy creators must constantly track market conditions to identify opportunities, evaluate new protocols, and execute strategies.

AI has the potential to overcome these challenges. By leveraging machine learning and automation, DeFi platforms can now analyze large amounts of on-chain data, identify patterns, and execute strategies at a much more efficient level than human managers. Understanding user needs in natural language helps to increase its scale by making DeFi accessible to a large number of users.

A game-changing areas

DeFi offers unparalleled selectivity, but remains difficult to use. CEX is simple to use, but limits user control and selectivity. AI provides the opportunity to bridge this gap. By automating complex DeFi interacts and simplifies the decision-making process, and AI can make DeFi as easy to use as a centralized platform without sacrificing selectivity. On the other hand, AI can help CEX make a new listing decision faster, providing more options than it is currently available.

A practical example of this is Hey Anon, an AI-powered DeFi interface. I personally tried Hey Anon; It is efficient in exchange and cross-chain, without manually finding contract addresses or selecting bridges. The entire interaction process is chat-based, which makes it easier for new users to use. However, it is slower than manually executing these transactions. Additionally, it currently lacks support for manual transfers—a important feature that should be incorporated to provide more flexibility.

Is there a market for DeFi + AI?

Before exploring the intersection of artificial intelligence and decentralized finance, let's take a step back and look at the total addressable market (TAM).

As of active and passively managed regulated open-end funds exceeded $80 trillion in asset management (AUM) as of the third quarter of 2024. By comparison, the total assets managed by Bitcoin (BTC) and Ethereum (ETH) ETFs as of January 21, 2025 were $150 billion.

These data highlight a key point: trillions of dollars are managed by professionals around the world because most people prefer not to deal with their finances directly. They tend to choose products that are easy to use and provide stable growth. Cryptocurrencies should not be different either. We have seen this from the phenomenon that users prefer centralized exchanges (CEXs).

The volume of centralized exchanges is still about five times that of decentralized exchanges. An important factor in this gap is availability. Managing wallets, navigating contract addresses, and understanding on-chain processes is difficult for many people. But it also brings huge benefits. Perhaps the biggest benefit is the possibility of early profit. If you find TRUMP on the chain, it has a market capitalization of less than 10 $100 million, then when it goes live on a centralized exchange, you have already earned five to ten times the gain. This is increasingly applicable in the market players to players (i.e., PvP) stage, where net inflows stagnate. Assets are exchanged between existing participants.

Rotation is the name of the game. There is a new hot flavor every week.

Even if you have been in the cryptocurrency space for a long time, it will be hard to catch Jailstool or CAR Opportunities. You only have one day to understand it, do due diligence, buy and sell – for most people it’s nearly impossible to complete without prior knowledge. The only way to seize this opportunity is to design a system that combines on-chain metrics such as newly deployed contracts with surges in trading volume and price and social media messages like X. Both tokens are currently down more than 80% from their respective highs and have not yet been listed on any major centralized exchanges.

The round of price discovery has ended. A large number of trading activities have been conducted on DEX and/or OTC tables. Early participants such as traders, liquidity providers or arbitrage players have established an informal market price. When the asset reaches the CEX, most of the initial volatility and price exploration have already occurred.

In addition, most centralized exchanges charge higher redemption fees compared to venues such as Jupiter and Raydium. Jupiter does not charge any fees, while Raydium charges a 0.25% fee for each redemption. The Moonshot trading app charges a 2.5% fee to users, while exchanges such as Binance and Coinbase charge different fees based on the user's trading volume. These fees usually range from 0.1% to 0.6%. One model can be seen from these fees--platforms with better user experience can charge higher fees.

Coinbase has more than 110 million users, far exceeding the active user base of DeFi. Given this huge gap, the potential total available market size of DeFi is huge. Even if it is not billions, it is conservatively estimated that DeFi can strive to attract a considerable number of current centralized exchange users, provided that it can do a good job in usability. This is where AI can play a transformative role.

Enter DeFAI: Simplify DeFi with AI

DeFAI, an emerging DeFi trend aimed at simplifying the DeFi user experience. It will be as simple as talking to a broker to buy and sell stocks – just better. You will interact with an AI agent that can convert text or speech into deterministic on-chain operations and provide you with suggestions powered by data.

So, when a token is published on a chain that you are not familiar with or have never been bridged to, you can go to the chat interface and tell the AI ​​that you want to bridge the assets to this new chain to perform an XYZ operation. The AI ​​agent will complete the process for you.

We wrote in our article on chain abstraction and smart wallets that both are tools to enhance the user experience of cryptocurrency. Chain abstraction eliminates the complexity of managing chains and bridging, while smart wallets utilize technologies such as pass keysTo simplify and protect wallet management.

However, AI agents have the potential to really expand DeFi’s cake. While progressive improvements have been made in addressing user experience challenges, AI agents can help DeFi cross the adoption gap if executed properly.

Visualization Impact

Now, DeFi's user base is composed of developers, advanced users, and late-chain adopters. As AI agents lower the entry threshold, the DeFi user circle can be significantly expanded to attract more CEX users who were originally willing to avoid the complexity of decentralized finance.

Abstract Meet Intelligence

Abstract user experience is just one of the things that artificial intelligence agents can help. Intelligence is the second aspect. Think about ordinary centralized exchange users. They are unlikely to already know about the on-chain applications that can be used and the assets that can be considered investing or trading. These contents must be planned for them. In the early days of the internet, Yahoo was a mastermind who helped millions of people discover and browse the web. App stores today play a similar role, deciding which apps get exposed and which won't.

Centralized exchanges have already acted as curator to some extent. What tokens they choose to list actually determines what most retail users can easily trade. If this curation feature is cancelled by forcing users to move to on-chain transactions, discovering opportunities and applications will become a difficult task. Users need a trusted wizard to guide them through this complexity. The question is: Will AI agents democratize this curation, or will they simply transfer power from centralized exchanges to the hands of those who control these agents?

The combination of curation and intelligence is the real power. It is not enough to just present opportunities; users need context, analysis, and execution strategies.

With so much happening on the chain, how can new users start evaluating opportunities? Many questions need to be answered. What applications do you use for lending and trading? Where to buy NFT? How to find the correct contract address? AI tools/agents like AIXBT can inform abstract tools like Wayfinder and Hey Anon.

AIXBTis a proxy that swallows information on X and puts it in context. It posts hundreds or even thousands of tweets every day. Sometimes its tweets or posts can even affect the market. Shlok wrote his paper on AIXBT. The paper notes that the agent stands out because of its deep integration into the crypto community, its complex analytical capabilities, and its potential to achieve growth through intellectual property and consumer engagement. The future of AIXBT may develop into a significant player in the AI ​​and crypto consumer market, provided it continues to innovate and maintains transparency in operations.

One ​​of the teams we have been working closely with is GudTech, who are committed to simplifying the entry process for retail users. GudTech was created by Zircuit's relevant team, and its vision is to provide contextual information while achieving transaction execution. Let me explain. As for the example of TRUMP token above, users may be unsure whether the US president has actually issued the token, or whether there are multiple well-known big wallets buying the token in large quantities. You may just buy it directly after seeing the token code on DEX without enough background information. One of the biggest problems in the current cryptocurrency space is that there are already 34 million tokens (and the number is still increasing), but there is little relevant context information. The crypto space is full of unstructured and fragmented data that is often biased and unreliable.

Gud combines on-chain data and contextual information of social networks to allow for direct purchase of assets on-chain. It solves the problem of reducing the learning curve and cognitive burden for new users to enter the crypto field. You could have seen the asset rise 100 times in the past 24 hours, and President Trump did post the stock code on Twitter.

In an ideal world, Gud even validates the contract address and executes transactions for you. Gud is building an agency economy, through a dialogue interface, where users can purchase all assets on the chain and get context information from the perspective of cryptocurrency native users. Gud terminals also have critical thinking ability and are able to reason about the positive or negative aspects of transactions. In addition, Gud terminals can use up to 10 queries for free every day, similar to Web2 platforms such as Perplexity, focusing on incentivizing adoption and use rather than hoarding tokens.

This future may seem a bit distant, but this model is mainly based on two aspects. First, it is how to capture, contextualize information and share it with newcomers in the industry. Imagine a private wealth manager explaining the latest trends in the industry to you. This is already in the field of consulting or lawIf that happens, you can get 80% insight by launching a ChatGPT instance.

There is currently no environment required for such interactions that meet the cryptographic native needs. Gud aims to package it into a simple experience to expand the number of users in the current crypto space. However, they are still in progress. As of this writing, the product’s trading system has not yet been launched, and there have been several wrong interactions on the agents on Twitter. But we will eventually achieve this.

Wayfinder is another highly anticipated application developed by the same team that built Parallel, a leading blockchain game. Here is a demonstration showing how the Wayfinder agent aggregates funds from multiple chains and sends them to different wallets. Hey Anon has integrated multiple chains and applications. It combines the ability to execute transactions with real-time insights from multiple platforms including Twitter, Telegram, and Discord.

Imagine: you open a sophisticated interface similar to ChatGPT or Claude and start talking to your personal AI transaction agent. You share your risk tolerance, investment goals and preferences. The agent understands your parameters and manages your portfolio automatically – executes trading, opens positions and adjusts strategies in real time within the boundaries you define. This is not science fiction; this is where we are heading. Here are a glimpse of what is possible.

Reality test

Applications like WayFinder are not available to everyone. But before being attracted by the hype and token prices brought by the DeFAI narrative, it is crucial to take a step back and evaluate reality. The sobering fact is that we haven't reached that level yet. I don't fully understand the engineering complexity required to achieve our goals, so I can't predict how long it will take. But it is obvious that there are still significant gaps in intelligence and abstraction in DeFi that need to be filled.

For example, taking AIXBT as an example, it can be said to be the best intelligent or information synthesis agent in the field. It generates multiple tweets every day, making it impossible to manually evaluate each investment or trading idea. If you follow all its advice in the $10 million to $100 million range, you will get an average return of 2% with a 39% win rate. This shows that while AI can process large amounts of data and discover opportunities, it still lacksThe meticulous judgment of inexperienced traders. In addition, there is another important note for this performance: a few tokens perform significantly better than other tokens. If you miss those few winners, you’re likely to suffer from AIXBT’s advice.

With this warning, it is easy to ignore the value of AIXBT. But this is related to a long-standing debate in traditional finance: Is active investment really better than passive investment? "Walking on Wall Street" popularized the view that the market is largely effective, and it is difficult for even professionals to continue to beat index funds. In fact, studies have shown that the returns generated by randomly throwing darts into stock lists can be comparable to those of professional investors. This highlights a broader reality that markets are unpredictable and human expertise alone does not always guarantee advantages. However, the performance of The Medallion Fund's continued victory over the market for 30 consecutive years proves that when human intelligence is combined with algorithms, it can indeed create advantages.

I personally cannot keep up with AIXBT's tweets to make transaction decisions. However, I would use a filter to distil thousands of AIXBT tweets into the top five trading ideas. Currently, it serves as a good filter, but requires significant optimization. An additional dimension is needed to be added to it—a dimension that effectively filters its output and makes smarter and more strategic decisions. The challenge of intelligence lies not only in quantity; it also in priority. What we need is a complex filtering system that can refine AIXBT's numerous suggestions into actionable, high-probability transactions.

The current shortcomings of AI

Recalling the intelligence issue, I want to understand how the execution/abstract work. I tried using Orbit to buy the meme coins that it considers the greatest potential. I interacted with "Meme_Radar_TK_Agent" but didn't get the result I wanted. I had to repeatedly clarify my requests with the agent. Although I chose the AI-recommended token, it failed to retrieve relevant information about the token. The agent has difficulty with basic tasks: it will recommend a token, but then cannot provide key details about its own suggestions.

Orbit ($GRIFT) in 1The transaction volume reached $180 million on the 22nd of the month. However, it cannot perform a simple task smoothly for first-time users. This reveals a significant gap between the analytical capabilities of AI and its ability to efficiently execute real-world transactions.

Of course, this category is still in its infancy and the product will evolve over time. Our own product, SentientMarketCap, is under public development and is constantly improving based on user feedback and actual testing.

Similarly, platforms like Griffain and WayFinder may provide enhanced solutions, but they are still largely untested in real-world environments. The entire DeFAI field remains an evolving experiment, with products actively improving through continuous iteration and real-world insights.

It is obvious that a successful DeFAI platform needs to perform well in three key areas:

1. Reliable intelligent systems that continuously organize contextual data to identify profitable opportunities 2. Seamless execution, minimizing friction between decision-making and action 3. User-friendly interfaces that enable ordinary users to easily perform complex DeFi operations

Technology is rapidly evolving, but we are still in the early stages of this evolution. The key is to manage expectations while continuing to innovate and improve these systems based on actual performance and user feedback.

The application of artificial intelligence in decentralized finance is not without risks. Undertrained models, dependence on historical market conditions, and potential possibilities for manipulation are all issues that need to be addressed before AI-driven decentralized financial platforms achieve mass adoption.

Learn from Feynman

Richard Feynman: Can Machines Think?

https://youtu.be/ipRvjS7q1DI

Richard Feynman's argument on machine intelligence is highly correlated with DeFAI. He believes that machines can do better than humans on specific tasks. If we can combine these specific tasks into a superset—a new system—it can significantly help us make decisions and execute in financial markets. AI in DeFi should follow this principle: it should not replace human intuition, but should enhance our capabilities to create seamless experiences for users by integrating multiple intelligent dimensions—automatic execution, market analysis and risk assessment.

This modular approach to artificial intelligence capabilities has a profound impact on the development of DeFi. DeFi requires not only automation, but also intelligence that can optimize execution. Take a well-managed hedge fund as an example. It has different teams, each with expertise in a specific field. Some teams focus on executing transactions with minimal slippage, others analyze patterns to predict market trends, while the third team ensures that funds flow efficiently between different markets.

AI agents can operate in the same way in DeFi. A proxy can execute transactions efficiently by reducing price impact and avoiding MEV attacks. Another can detect patterns in the on-chain data to predict liquidity changes or market trends. For example, this agent can access tools such as GMGN and Cielo, tracking wallets on-chain to assist with other analytics. The third one can manage cross-chain transfers to ensure that funds are best allocated in each ecosystem. When these agents are combined, they go beyond simple automation. They bring intelligence to execution—from providing transaction input to ensuring transactions are conducted at the best price, minimizing risks, and seamlessly across multiple networks.

Towards proxy collaboration

Most DeFAI products are trying to solve the problems of intelligence (analysis, synthesis) and abstract (execution) capabilities, and there is good reason. Any single component offers limited value, just like having maps without vehicles or vice versa. But the real power lies in specialization and integration.

The current landscape is similar to a decentralized ecosystem, with different agents performing well in different fields. Some agents excel in market analysis and pattern recognition, while others excel at performing complex DeFitrade. The best solution may involve cooperation among agents, leveraging each other’s strengths. Imagine Anon’s expertise in DeFi integration combined with AIXBT’s analytical capabilities—a collaboration that creates a seamless experience that translates market insights into executed transactions.

Listen is moving in this direction. The idea is to create a system that allows multiple AI agents with dedicated capabilities to collaborate on the complexity of DeFi. By integrating these agents, it aims to automate not only individual tasks, but end-to-end financial strategy automation.

This approach will allow users to issue complex commands through a simple conversation interface (voice and text), such as portfolio rebalancing or income farming across multiple protocols, making once a difficult task for experienced DeFi users easy to manage for ordinary people. The collaboration with Arc aims to empower these AI agents by providing a platform to interact, learn, and scale. This ensures that the execution layer and the smart layer are not only separate, but also work together to provide a comprehensive DeFi experience.

Familiar evolution

The current state of DeFAI is reminiscent of the early banking industry. Initially, financial services were decentralized—users had to visit different institutions to pay bills, make investments, and transfer money. With the launch of banks, integrated platforms emerge to provide seamless financial management in one place.

DeFAI needs to usher in its own "super application" moment - a platform that can seamlessly integrate various professional agents. It can be considered as a coordinated system where the analytical agent provides market intelligence, executes the agent to process transactions, risk management agents to monitor positions, and portfolio optimization agents to balance asset allocation.

This integration will create a unified experience where users interact with just one interface, while multiple specialized agents work together behind the scenes, just like modern food delivery applications deal with everything from restaurant discovery to payment processing. The future of DeFAI lies in creating ways that allow professional agents to collaborate smoothly. This approach will allow each agent to focus on its core strengths while participating in a larger, stronger ecosystem.

Low the threshold and unlock the potential for adoption

Robinhood revolutionized retail investment by making stock trading accessible to millions of people who never considered participating in the market.After the outbreak of the COVID-19 pandemic, Robinhood added more than 3 million new accounts with funding in the first four months of 2020 alone. Of these, 1.5 million are first-time investors. This unprecedented growth is not only driven by zero commission transactions and intuitive mobile-first design, but also driven by external factors such as stay-at-home orders during the pandemic.

DeFAI has similar opportunities. The complexity of DeFi has long been a major obstacle to widespread adoption. The cumbersome wallet setup, confusing interfaces, and decentralized liquidity across multiple chains are all deterring those who are the most focused users. If DeFAI is to thrive, it must follow Robinhood’s approach – eliminating friction and making DeFi as simple as opening an app, selecting an asset, and executing a transaction in seconds.

Apart from usability, AI-driven planning may redefine the discovery process in the DeFi field. Just as Yahoo once planned early networks and app stores are now guiding mobile discoveries, I'm curious about what new business models will emerge around DeFi planning, driven by artificial intelligence. One open question is whether these innovations will empower users or simply transfer control from centralized exchanges to those who build and manage these AI systems.

Our AI applications in the DeFi field are still in their early stages. The next few years will determine whether these technologies have truly achieved the democratization of decentralized finance or whether they have paradoxically introduced a new form of gatekeeper. The challenge is not just about automation—it is about ensuring AI enhances accessibility, transparency and decentralization, rather than replacing the original gatekeeper with another set of gatekeepers.

Waiting to use the new era of DeFAI.