When we hear about the intersection of AI and crypto, the projects that have gained major traction in the category are the networks that are solving for data collection, GPU computation or data inference in the AI space. These include protocols like Akash Network, Ritual net, etc. They differentiate themselves in the big AI industry by leveraging decentralisation, incentivisation, censorship resistance and privacy that web3 provides.

‍

While these projects are creating fascinating applications, their impact on everyday web3 users remains limited, and they're not effectively bringing new users into the web3 space.

‍
The rise of AI agents

With the rapid growth of Web3, new crypto protocols, tokens, and apps are popping up all the time. Navigating this complexity can be daunting for even the most experienced users. Hence there's a growing trend of creating AI Agents, which are smart helpers that make using crypto apps easier. These AI Agents sit at the crossroads of crypto and AI, aiming to fix the tricky user experience (UX) issues in crypto. Imagine a future where you could simply tell an AI agent what you want to achieve on-chain, and it would automatically compose and execute the necessary transactions for you.

‍

AI Agents will help us build an intelligence layer on top of the existing DeFi rails. This layer will become the TradFi equivalent of bankers, investors, traders, fund managers etc in web3. And they will leverage the underlying primitives to transact on-chain.

DeFi-AI integrations promise to introduce advanced applications like AI-driven lending, intelligent yield farming strategies, automated market making, and AI-assisted portfolio management. The usecases don’t end there, AI Agents can also be leveraged in the gaming industry, by allowing users to have assistants and pets that can help them navigate through the game and can enhance the overall user experience of the game.

Based on the type of use case and value add the protocols bring, the evolving AI Agent space can be largely broken down in the following categories:

‍

Caterogies of AI agents

Gaming AI Agents:

Teams like Parallel Colony are building gaming AI Agents that enhance the user's gaming experience. These AI agents operate within Web3 environments, interacting with players and game elements through on-chain smart contracts. The AI Agents are designed to either act on the behalf of the user or as an in game pet/assistant. These agents are being designed such that they can also interact and trade assets with other agents.

‍

There are also games both web2 and web3 who are actively using AI to design dynamic non-playable characters (NPCs) in the game. But for the purpose of the article we will only focus on the AI Agents that can trade and act on users behalf.

‍

Autonomous Portfolio Managers:

These are AI Agents that can be leveraged to manage a pool of assets from multiple different user profiles. The aim of the AI Agent will be to maximise the pool’s earnings by delegating the assets across various DeFi strategies by leveraging the right set of off-chain AI data streams. It's essentially a portfolio management service leveraging the power of AI. To make sure that the protocol is trust minimised, projects are also enabling ZK (via protocols like Modulus) to provide the on-chain proofs of the AI inferences they generate in the process.

‍

Prompts based AI Agents:

Imagine a future where you could simply tell an AI agent what you want to achieve on-chain, and it would automatically compose and execute the necessary transactions for you.

This is what majority of AI Agent projects are thriving to build, and we can imagine a future where prompts might become the preferred way for general users to interact with the blockchains.

‍

Projects like Wayfinder, Brian Knows, Aperture Finance and others are working on developing ChatGPT-like interfaces that can help users make knowledgeable transactions on the blockchain directly by chatting with the AI Agent. These protocols leverage the power of LLMs to convert user’s prompts and intents to an executable transaction.

‍

Let's discuss a few of these AI Agent protocols in detail:

Autonolas Agents

Autonolas is a platform that enables the creation and management of autonomous agent services. These services, known as agent services, operate independently off-chain as a multi-agent system (MAS), collaborating to achieve shared objectives. In essence, Autonolas empowers developers to build and deploy autonomous agents that work together seamlessly off-chain while leveraging blockchain technology for enhanced capabilities on-chain. One such agent is BabyDegen.

‍

Here is a directory of other agents built on Olas Network

‍

AutoTX by Polywrap

Polywrap is building a network of specialized AI agents that perform complex tasks for web3 users and protocols. These agents leverage crowdsourced insights, off and on-chain data sources, task planning, and batch transactions to efficiently solve problems and make decisions across various domains. Current agents include those for payments, market research and trading, social content curation, predictions, and public goods funding. Future plans for Polywrap involve expanding the range of specialized agents, decentralizing their execution, and evolving the system through community-driven governance. One such AI Agent is AutoTx.

‍

AutoTx can translate high-level user goals into a series of blockchain transactions. This means you no longer need to be an expert in every protocol or spend hours learning how to manually compose different types of transactions. Simply tell AutoTx what you want to achieve, and it will handle the rest.

‍

‍

Parallel Colony

Parallel Studios is taking a fresh approach to AI agents with Colony, a new AI-powered Web3 survival game. In Colony, highly autonomous AI agents, or "avatars," continuously learn from their environment. Players must guide and collaborate with these avatars, which have diverse skills and abilities, to survive in a future Earth with competing colonies.

Colony stands out by integrating continuous learning into its gameplay. The AI avatars develop unique personalities and worldviews, learning from their own experiences, identities, and goals. Additionally, these avatars can autonomously manage digital assets through dedicated Web3 wallets, allowing them to trade with other in-game avatars.

Read Whitepaper

‍

Wayfinder

Wayfinder is creating a "map" for AI agents to handle tasks and simplify on-chain activities for users. By open-sourcing development and incentivizing builders with the $PROMPT token, Wayfinder will expand a network of navigational instructions. Wayfinder paths will continuously enhance AI agents' capabilities, making them smarter over time. It aims to connect blockchains and off-chain data sources, allowing users to execute tasks easily via command prompts. Their innovation aims to make blockchain interactions more efficient and accessible, improving users' lives by reducing complexity and stress. You’ll love this analogy and explanation by @tiggity_tc on Wayfinder.

This is a video of the wayfinder agent in action

‍

Whitepaper

‍

Noya

NOYA is a decentralized finance (DeFi) protocol that empowers AI agents to manage liquidity across multiple blockchains with security and precision. It uses a composable system built from the ground up, including a private keeper network, an AI-compatible oracle, and a competitive environment for AI and strategy managers. Noya has multiple vaults, each catering to a separate user intent profile. The protocol has its own designed AI Oracles to read various DeFi markets and passes information to the AI Agent.

NOYA's infrastructure supports multiple intents like liquidity provisioning, leverage management, and borrowing rate optimization, using advanced technologies such as Zero-Knowledge Machine Learning (ZKML). It aims to set a new standard for omnichain liquidity management and financial strategies. The team is closely rolling out access to the protocol.

‍

Brian Knows

Brian offers APIs which developers can integrate into their applications to provide users with the ability to generate web3 transactions by passing in their intents via prompts like "Can you swap 10 usdc for eth on Uniswap on Ethereum mainnet?" They also provide smart contract deployment services via prompts. On the backend, team uses LLMs to convert the prompts into a web3 transaction and then get it executed via their preferred protocol and solver integrations.

‍

The team has also built out a Brian app that you can use to explore the feature set. The team is looking to expand their offerings by providing recurring and automated payment setup optionalities to their users, and much more.

‍

Aperture Finance

Aperture Finance revolutionises DeFi by offering liquidity management services through a user-friendly protocol. It aims to enhance the DeFi user experience with an intuitive chatbox interface inspired by GPT, allowing users to express their goals in natural language. Third-party participants, known as Solvers, handle requests by optimising processes to ensure efficient and cost-effective execution.

‍

‍

Fungi Agent

Fungi leverages the power of smart accounts and account abstraction to provide a self-custodial AI Agent experience . Fungi lets users instruct command prompts through its interface, it then processes real-time blockchain data and autonomously executes actions based on user’s instructions.

‍

Users can chat with Fungi to deepen their understanding of crypto, receive personalized guidance, perform on-chain transactions, create customized DeFi strategies (Hyphas), and even monetize these Hyphas by sharing them with the community. Fungi works as a network of agents who interact with each other and learn from past experiences—a financial superintelligence accessible to everyone.

‍

Here is how Fungi works:

Source: https://words.odisealabs.com/defi-ai

‍

Fyde Protocol

Fyde enables users to grow their crypto holdings faster by depositing into diversified, AI-managed vaults that lock in gains and reallocate assets based on market performances and reduced volatility.

‍

Users can deposit various tokens into these vaults and receive $TRSY, a token representing their share of the vault's assets. Fyde aims to maintain consistent $TRSY liquidity across market conditions, making it easy for users to transact.

‍

The need for an AI auth Layer for smart agents

In all of these upcoming AI & intent related projects, the potential use-cases range from abstracting simple tasks to empowering AI agents to execute complex DeFi strategies to find you the best yield. However there are two main challenges that these AI agents face:

‍

• They can’t be truly autonomous: Currently AI agents can recommend on-chain actions and prepare transactions for the user, but still need the user to sign these or approve.
• If they choose automation, they lose out on security: Protocols tend to explore alternate solutions for automations, things like approvals, centralised vaults, shared private key pairs, etc that make protocols the custodials of your assets and introduces significant risk.

‍

We need guardrails for the AI in the form of user-defined permissions - strictly defining what actions the AI is allowed to sign & which it does not have the authority to. Thus, we need a solution which can delegate transaction authorisation to AI agents, but only within specific permissions & rules.

‍

‍

We’re solving for this at Biconomy. If you’re building the next generation of autonomous agents, do reach out to us - we’d love to chat!

‍

Interesting articles on AI Agents for reference:

AgentFi: The Real Case for Crypto x AI

Bridging the Gap: AI as a Catalyst for DeFi Adoption

Infinite Worlds

Understanding the Intersection of Crypto and AI

Crypto AI Agents: The First-Class Citizens of Onchain Economies

The Internet of Agents

The Convergence of Ai, DeFi and Intents with Aperture Finance

------------------------------------------------------

This piece was authored by Nishil jain