At Talus, we’re building the infrastructure for a Smart Agent Revolution.

Smart agents are AI actors that are capable of carrying out real world actions on a human’s behalf toward a human’s goals. Here’s an example:

If you ask ChatGPT to plan your trip to Paris, you’ll get a list of hotels, flight prices and dates, sights to see, and restaurants to try. If you tell a smart agent that you want an incredible vacation in Paris, it will buy the cheapest flight, book your hotel room, plan your itinerary, make restaurant recommendations, and everything else you need to achieve your goal.

Onchain smart agents unlock that same power, but with the many benefits of blockchains. At Talus, we’re laying the foundation for onchain smart agents that are powerful, decentralized, and secure. Smart agents will not only perform much of what is currently possible on blockchains for us, but they will expand the frontiers of web3 themselves.

Let’s explore just some of the most exciting use cases for the Talus onchain smart agents.

Web3

1. Portfolio Management

Smart agents could perform intelligent portfolio management for you, with far less cost, effort, and required trust than a human. Instruct the smart agent on your investment goals and risk tolerance, and it will make calculated decisions to rebalance your portfolio, optimize and diversify your investments, and identify new opportunities.

2. Intelligent Oracles

As more real world assets are tokenized and brought onchain, smart agents can be used as intelligent oracles to transmit price data from offchain. This is especially valuable for illiquid and difficult-to-price commodities, such as watches, trading cards, or real estate. An onchain smart agent can automatically analyze the market to determine valuations in a decentralized, trustless manner.

3. Natural Language UX

AI models can use Natural Language Processing to understand user intents, and intelligently execute them based on context, market analysis, user goals, and risk aversion. These could range from simple asks, like “Buy $100 of Bitcoin at its best price this week,” to complex tasks like “Rebalance my portfolio to take advantage of the memecoin bull run.”

4. AI-Enhanced Insurance

Smart agents could automate a vast swath of the insurance business. These include managing policies in open smart contracts, collecting premiums and distributing payments from decentralized vaults, evaluating claims, detecting fraud, and actuarial calculations.

5. Smart Wallets

Smart agents can be built into wallets to make them more powerful and autonomous. Your wallet could understand activity to automate your life, making and requesting payments as needed. A smart agent could actively, intelligently monitor for fraud, reject incoming scams, and identify suspicious activity. The wallet could even perform onboarding tasks and connect your wallet for you, making UX faster and smoother.

6. Decentralized Credit Scoring

Smart agents with access to onchain data can assess creditworthiness accurately and fairly. They can also adjust rates and payment schedules in real time in response to changing balances and financial activity. The result is a transparent, verifiable system in response to today’s hierarchical, centralized, and opaque one.

DeFi

7. Liquidity Management

In contrast to today’s arbitrage-based DEX liquidity strategies, decentralized smart agents could manage liquidity to lower slippage, fees, and impermanent loss. An embedded smart agent in each pool could monitor market conditions to predict volatility and react to activity, then dynamically reallocate capital.

8. Onchain Managed Funds

Dramatically lower fund fees with smart agent managers. They can allocate assets, monitor markets, perform sentiment analysis, and read whitepapers and reports. Smart agents could even be trained to become experts in specific sectors, markets, and projects. The result is wider access to more powerful investment strategies.

9. Improved MEV

Smart agents embedded in order books and mempools will reimagine how transactions are ordered and processed. AI can intelligently predict future block value, which is far more powerful than today’s simulation-based methods. This enables MEV strategies that maximize value for the ecosystem, while minimizing exploitation and negative outcomes for individuals.

10. Transaction Optimization

Smart agents can automate transaction routing in ways that are far more dynamic than existing bots. User- or DEX-owned smart agents could analyze swap paths in real-time to optimize for lower fees and higher returns for both individuals and ecosystems.

11. Intelligent Debt

Smart agents could improve every aspect of the credit market, from loan underwriting to bond issuance to risk assessment. AI could perform credit worthiness checks, responsively adjust interest rates and repayment schedules, and collect automatically with smart contracts.

DAOs

12. Treasury Management

Trustlessly manage pooled funds with smart agent treasury managers. Intelligent asset and budget management, automated collection and distribution through open smart contracts, and AI-powered decisionmaking are all possible with AI. This could dramatically improve transparency and efficiency, particularly in groups with a large proportion of anonymous members.

13. Smart Delegates

Voter fatigue and apathy are challenges for DAOs and DAO members alike. Smart agents could be given your personal goals and goals for the organization, then empowered to analyze proposals, predict outcomes, and vote accordingly. They could even use NLP to generate arguments and participate in debates on your behalf.

Security

14. Content Moderation

Smart agents could perform the tedious work of content moderation for decentralized social networks. They could tirelessly monitor for AI-generated content, ToS violations, spam, scams, and abuse. The combination of onchain posts, accounts, and moderation means a fairer and more open future for the read-write Internet.

15. Identity Verification

Smart agents could rapidly analyze new accounts and connected wallets to identify synthetic profiles. Onchain data could be used to identify malicious actors and fraudulent activity, protecting projects, ecosystems, and other users.

16. Constant Auditing

Auditing is currently an underperformed, time-consuming, and expensive process prone to human error. Smart agents trained on software principles and project goals could perform constant monitoring of smart contracts and code, to detect new weaknesses and unexpected changes. They could even react to new exploits in other projects to identify and repair holes in response.

Decentralized Gaming

17. Enhanced Prediction Markets

Smart agents can perform the work of skilled researchers to create the most accurate real-time odds. They could also manage liquidity, to ensure smooth operation even in niche markets. Fraud detecting smart agents could compare onchain markets and offchain events to identify tampering or interference. A smart agent trained on your goals could even be your “gambling assistant,” pointing out the best odds, parlay and hedge strategies, or even playing on your behalf.

18. Playtesting

A crucial aspect of game development is playtesting, in which players demo the game to identify flaws. This process could be automated, sped up, and made less expensive with smart agent players. They could even be customized to represent various skill levels, language fluencies, and incentives to make better games.

19. AI NPCs and Opponents

Instead of stiff, robotic bots, NPCs and computer opponents will be replaced with smart agents that act dynamically and humanistically in the game. The result is more fun gameplay and more immersive worlds.

We’ve just barely scratched the surface of the potential of onchain smart agents. As we step forward to the Smart Agent Revolution, the possibilities are endless. What smart agent use cases can you think of?

About Talus Network

Talus is building AI Smart Agents for the people.

We are creating a decentralized protocol where individuals can freely exchange data, models, and computational power to make AI transparent and verifiable. By leveraging the security, performance, and developer experience of Move, Talus is the foundation of the Smart Agent Revolution and a new era of consumer AI.

To learn more about Talus, follow us on Twitter, check out our GitHub, and join our Discord.