Fetch.ai employs a unique consensus mechanism called UPoW. This system allows even less powerful nodes to earn a share of the block rewards. UPoW works by generating new blocks similar to standard Proof-of-Stake protocols, but it also ranks computing problems by difficulty and packages them into proof-of-work packages.

Fetch.ai’s platform uniquely combines blockchain, AI, and machine learning to create a decentralized digital economy. Its architecture and core technologies foster an environment where Autonomous Economic Agents (AEAs) thrive, facilitating complex, automated tasks across various sectors.

The foundational structure of the Fetch.ai network includes the Almanac contract, the Fetch Name Service, and the Fetch Ledger. The Almanac contract is a repository storing critical information about AI Agents, while the Fetch Name Service provides human-readable naming conventions for resource identification. As the backbone of the decentralized infrastructure, the Fetch Ledger ensures secure data sharing and transactional integrity.

Autonomous Economic Agents (AEAs)

AEAs are software programs that operate autonomously in the Fetch.ai ecosystem. They communicate, negotiate, and collaborate using a common language known as Agent Communication Language (ACL). These agents can represent individuals, businesses, or devices, and handle various tasks. AEAs are created using the AEA Framework, a Python-based toolkit that equips developers with the necessary tools for developing and testing these agents. They can be managed and deployed using the AEA Manager, a web interface offering access to various services.

• Autonomy and Economic Focus: AEAs operate without constant oversight from their owners, focusing primarily on creating economic value. They are designed to handle various tasks autonomously, with minimal or no intervention.
• Cross-Chain Compatibility: The AEA framework supports the development of applications that can interact with different blockchain networks like Ethereum, making Fetch.ai’s platform versatile in the blockchain ecosystem.

• Task Efficiency: AEAs have tools within the framework to navigate the network, execute operations efficiently, and pursue new objectives.
  The AEA framework is an all-encompassing development suite for creating and managing AEAs. It provides the necessary tools for efficient agent operation and enables the developers to give AEAs specific sets of inputs for various situations. The framework includes:

• Simple Open Economic Framework (sOEF): Enables AEAs to navigate the Fetch.ai network, execute tasks, and search for goals.

• Agent Communication Network (ACN): Facilitates information exchange and communication among agents.
• Connections, Protocols, and Contracts: These components define agent-to-agent and component-to-component interactions, wrapping access to smart contracts and other services.

AEAs in Fetch.ai can be applied in numerous domains, including but not limited to:

• Automated Blockchain Interactions: Automating user interactions with blockchains, such as transactions or smart contract operations.
• Data Queries and Server Interactions: AEAs can interact with servers as clients or maintain servers to serve other clients.
• Decentralized Finance (DeFi): Automating complex operations in the DeFi space, like trading and asset management.

AI Agents

The concept of AI Agents introduced by Fetch.ai through the uAgents Framework is a fascinating development in decentralized systems and artificial intelligence. At its core, the framework aims to facilitate the creation of autonomous, intelligent software programs capable of interacting and collaborating with one another in a decentralized network.

One of the key advantages of this approach is its potential to revolutionize various industries by enabling AI Agents to tackle complex challenges, streamline processes, and improve decision-making. The documentation provides compelling examples, such as the application of AI Agents in supply chain management and financial services. Imagine a supply chain where AI Agents seamlessly communicate and coordinate, optimizing everything from demand forecasting and inventory control to logistics and risk mitigation. This level of intelligent automation could significantly enhance efficiency, reduce costs, and provide real-time visibility across the entire operation.

Similarly, in the financial sector, AI Agents could automate trading transactions, perform sophisticated risk assessments, detect fraud patterns, and even provide personalized financial advice to clients based on their profiles and market conditions. The potential for AI Agents to simplify processes, uncover valuable insights, and strengthen security measures in this domain is immense. What truly sets the uAgents Framework apart, however, is its emphasis on decentralization and user control. By allowing developers to create AI Agents as either public or private entities, the framework strikes a delicate balance between transparency and privacy. This flexibility is crucial in sectors where confidentiality and data protection are paramount, such as finance or healthcare.

Public agents, with their protocol manifests and digests exposed, enable seamless collaboration and interoperability. Imagine a logistics company’s public AI Agent sharing its communication methods, allowing other agents to interact and coordinate transportation tasks directly. This level of openness could foster innovation and facilitate cross-industry partnerships. Conversely, private agents keep their protocol details hidden, ensuring a high degree of confidentiality and security. A financial institution, for instance, could develop a private AI Agent to handle sensitive transactions, shielding its protocols from external scrutiny while still being discoverable through the Almanac.

This duality of public and private agents aligns perfectly with the decentralization principles and trustless interactions underpinning blockchain and distributed ledger technologies. By empowering users to control the level of transparency and exposure for their AI Agents, the uAgents Framework embraces the ethos of decentralized systems while simultaneously addressing practical concerns around privacy and data protection. Furthermore, the framework’s open-source nature and its integration with Fetch.ai’s AI Engine and Agentverse platform create an ecosystem ripe for innovation and collaboration. Developers from diverse backgrounds can contribute to the framework’s evolution, fostering a vibrant community dedicated to advancing the capabilities and applications of AI Agents in a decentralized context.

Agentverse

Agentverse is a cloud-based integrated development environment (IDE) that serves as a portal to the broader uAgents Framework and its vast array of use cases. This powerful framework facilitates the development of decentralized AI Agents, enabling them to communicate, negotiate, and collaborate seamlessly within a trustless, decentralized network. One of the key advantages of the Agentverse is its user-friendly approach. With a straightforward graphical user interface, the platform reduces the barriers to adoption of AI agent technology, making it accessible to developers of all skill levels. This is achieved through the provision of predefined agent templates and use cases, allowing users to create AI Agents quickly and easily, even with only basic Python knowledge.

The Agentverse: Explorer

The Agentverse Explorer is a crucial component of the platform, acting as an intuitive discovery tool for AI Agents registered within the Almanac contract. This decentralized directory houses detailed information about registered agents, including their capabilities, protocols, and communication methods. Through the Explorer, users can search for and connect with other agents, filtering them based on unique protocol digests or specific attributes.

Agents displayed in the Explorer can be categorized as Hosted, Local, or Mailbox agents, each serving distinct purposes. Hosted agents are those developed and deployed directly on the Agentverse, ensuring continuous uptime and real-time code editing capabilities. Local agents, on the other hand, are designed for testing purposes, pointing towards local endpoints. Lastly, Mailbox agents are registered within the Agentverse Mailroom, enabling them to send and receive messages continuously, even when offline.

The Agentverse: My Agents

The “My Agents” section of the Agentverse is where developers can truly bring their AI Agents to life. With a targeted 100% uptime, hosted agents are always kept online and up-to-date within the Almanac contract. The deployment process is streamlined, allowing users to have a new running agent deployed in just a few clicks, either from a blank script or based on pre-built templates. Scalability is a key consideration within the “My Agents” section, as the platform is designed to scale with the number of messages an agent receives automatically. This ensures that developers can focus on crafting the most innovative agent code without the hassle of continuously deploying and managing their agents’ infrastructure. Security is also a top priority, with agents running in secure, partitioned Python environments, safeguarding your code from malicious attempts to access your information. Additionally, the Agentverse offers the option to subscribe to ban lists, enabling you to ignore agents identified as bad actors easily.

The Agentverse: Mailroom / IoT Gateway

The Agentverse Mailroom is a powerful service that addresses the challenge of maintaining a constantly running server for agent communication. By setting up mailboxes for your agents, the Mailroom allows them to retrieve messages and take appropriate actions, even when they are offline. This service is particularly valuable in scenarios where running a server becomes a challenge, as it manages incoming messages sent by other agents and serves as a trusted intermediary. Developers can download these messages later without the need for a constantly running server, enabling seamless communication between agents registered in the Agentverse and local agents.

One of the standout features of the Agentverse is its integration with wallets and the Fetch.ai Ledger. AI Agents hosted on the platform have the capability to interact with the Fetch.ai Ledger, enabling them to send and receive transactions, query balances, interact with smart contracts, and more. This integration strengthens the trust system within the Agentverse and allows for the transparent querying of transactions performed by any agent.

Beyond the development environment, the Agentverse offers a suite of advanced features, including the “Services” functionality. This feature allows developers to combine agents into service groups, enabling them to offer services to users. For example, in the context of a hotel booking service, developers can create individual agents representing different aspects of the service, such as reception, room management, and concierge services. These agents can then be combined into a service group, providing users with a seamless and integrated experience through a single interface.

The Agentverse integrates with the DeltaV chat interface and the Fetch.ai AI Engine, enabling users to provide human input that guides the AI Engine in finding and interacting with relevant AI Agents registered on the network. This powerful combination allows users to request specific services, such as booking a hotel room in a particular city, and the AI Engine will identify and communicate with the appropriate agents to fulfill the request.

Almanac Contract

The Almanac contract plays a pivotal role in the Fetch.ai ecosystem, acting as a centralized hub that facilitates discoverability and communication among AI Agents. Its significance lies in its ability to bridge the gap between the decentralized nature of AI Agents and the need for a coordinated, accessible directory of their capabilities and protocols.

Almanac serves as a decentralized directory, akin to the Domain Name Service (DNS) of the internet. It houses detailed information about registered AI Agents, including their capabilities, protocols, and communication methods. This organized repository of agent data enables users to explore and interact with various agents through Agentverse Explorer, filtering them based on unique protocol digests.

The registration process within the Almanac is critical for AI Agents to establish their presence and enable remote interactions. Agents must periodically update their registration details within specific block limitations to maintain current and accurate information, ensuring data relevancy for users. This dynamic process ensures that the Almanac remains up-to-date, reflecting the ever-evolving landscape of AI Agents and their capabilities.

One of the key features of the Almanac is its role in facilitating the distinction between public and private AI Agents. This flexibility is made possible through the interplay between the Almanac and the protocols that govern agent interactions within the uAgents Framework. Protocols, in this context, establish the rules and message structures that enable AI Agents to comprehend each other’s communications and coordinate effectively for task execution. These protocols are integral to the smooth functioning of the decentralized network, ensuring that agents can collaborate and work towards common goals.

When an AI Agent is designated as public during its development, its protocol manifest and associated digests are published within the Almanac, making its endpoint available for communication with other agents. This transparency fosters collaboration and interoperability, as external agents can directly interact with the public agent, understanding its communication methods and capabilities. Conversely, private AI Agents maintain a higher degree of confidentiality by not disclosing their protocol manifest to the outside world. While their digests are still visible in the Almanac, allowing for discoverability, the details of their actual protocols remain hidden. This ensures that only agents explicitly aware of the private agent’s protocol can communicate with it, providing a controlled and secure communication environment when necessary.

This duality of public and private agents, facilitated by the Almanac and the underlying protocols, aligns with the principles of decentralization and trustless interactions that underpin the Fetch.ai ecosystem. It empowers developers and users to strike a balance between transparency and privacy, tailoring the level of exposure for their AI Agents based on their specific requirements and use cases. Furthermore, the Almanac enhances trust and integrity within the ecosystem by requiring agents to verify their address ownership through signature validation during the registration process. This verification step ensures the accuracy of the information stored in the Almanac, mitigating the risk of malicious or unauthorized agents attempting to manipulate the directory.

Open Economic Framework (OEF)

The OEF underpins the Fetch.ai network, providing a decentralized marketplace for agents to discover and interact with each other, exchange data, and access services. It is powered by the Fetch.ai Ledger, a blockchain network that facilitates value transfers and coordinates the agents’ activities.

Fetch Smart Ledger

Fetch.ai’s Smart Ledger is an integral part of its ecosystem, providing a unique combination of blockchain technology and a Directed Acyclic Graph (DAG). This hybrid structure enables efficient transaction processing and vast data handling, which is crucial for the operation of Autonomous Economic Agents (AEAs) and their complex activities. The Smart Ledger integrates traditional blockchain elements with DAG features, allowing it to handle high volumes of transactions and data. This setup ensures that the network can support communication and transactions involving millions of agents simultaneously.

Unlike traditional blockchains that follow a linear structure, the Fetch Smart Ledger assigns transactions to different lanes. This approach enhances the network’s capacity to handle a large number of transactions concurrently. The ledger not only records transactions but also processes a vast amount of data, decisions made by AEAs, and interactions within the network, all in real-time. This record-keeping ensures security and transparency of operations on the network.

The Smart Ledger plays a crucial role in the functionality of AEAs. It underpins their activities, such as booking flights or hotels, and other services by recording transactions and actions taken by AEAs. The FET tokens are used within this framework for transactions, including paying for services and rewarding AEAs for their contributions.

Fetch.ai’s approach to combining AI, machine learning, and blockchain technology differentiates it from traditional AI systems. By leveraging decentralized structures, it enhances transparency and reduces manipulation risks. The integration of AI functionalities in the Smart Ledger facilitates the handling of complex tasks and vast volumes of data, a feature uncommon in conventional AI systems.

The Fetch Smart Ledger, with its advanced capabilities, is suitable for various industries:

• Management of patient records, scheduling, and personalized treatment plans.
• Real-time tracking of goods and transparent transactions.
• Automated trading and secure transactions.
• Efficient trading of excess energy with transparency.
• Personalized shopping experiences and secure transactions.

Fetch.ai, being a relatively new technology, has seen early implementation in sectors like mobility and automotive, decentralized finance (DeFi), and smart cities. Its technology has been applied in decentralized networks for electric vehicle charging and commodities trading platforms like Mettalex.

Smart Contracts in Fetch.ai

Smart contracts in Fetch.ai are implemented through the CosmPy framework. Here’s a detailed breakdown of their nature, language support, deployment, and execution:

• Deployment: Using the LedgerContract object in CosmPy, developers can deploy smart contracts. They require the contract’s path and a wallet object for deployment.
• Interaction: Once deployed, interactions with the smart contract are facilitated in various ways, such as querying and setting state variables. The _address _method is used to obtain the contract’s network address.
• Querying State Variables: The contract’s state variables can be queried using the query method, providing a dictionary specifying the query information.
• Updating State Variables: The execute method is used to update state variables, followed by wait_to_complete() to ensure the execution finishes.
• Clearing State Variables: State variables can be cleared using the execute method with specific instructions.
  This process allows developers to create and manage dynamic smart contracts within the Fetch.ai ecosystem, leveraging blockchain’s capabilities for secure and efficient contract execution.

Collective Learning and AI Integration

Fetch.ai employs a decentralized machine-learning protocol called CoLearn. This protocol enables AEAs to share data and models without compromising privacy or security. CoLearn is implemented using CosmWasm, a high-performance smart contract language, and includes a marketplace for buying and selling data and models.

Interoperability

The Fetch.ai network boasts interoperability with other blockchains and networks. This is achieved through a cross-chain protocol based on the Cosmos SDK, enabling Fetch.ai agents to access data and services from platforms like Ethereum and BNB Chain. The platform also facilitates the swapping of FET tokens with other cryptocurrencies using bridges such as Axelar.

AI Engine

The AI Engine’s primary objective is to analyze, understand, and link human input with the most suitable AI Agents, facilitating natural language interactions. It achieves this by reading user input, converting it into actionable tasks, and selecting the appropriate AI Agents registered in the Agentverse to perform those tasks. The AI Engine’s strength lies in its ability to understand and adapt to different contexts. It examines trends and turns random inputs into meaningful insights by evaluating previous interactions. When faced with uncertainty, the AI Engine solicits feedback from the user to verify that its suggestions and recommendations align with the user’s end goal. This contextual understanding allows the AI Engine to actively anticipate user needs and adapt accordingly, creating a seamless and intuitive experience.

The AI Engine’s smart routing feature plays a crucial role in its effectiveness. This feature enables the AI Engine to carefully evaluate the available agents registered in the Almanac, taking into account their performance spectrum and past performance data. By considering these factors, the AI Engine ensures that tasks are assigned to agents with the appropriate expertise, maximizing the likelihood of successful task completion and enhancing the overall user experience.

DeltaV: The AI-Based Chat Interface

DeltaV acts as a front-end interface to the AI Engine, providing users with a simple and intuitive chat interface through which they can enter their requests. These requests are then translated by the AI Engine into a series of tasks to be performed by the appropriate AI Agents. DeltaV has been developed with the intent of connecting users to AI Agents in a seamless and user-friendly manner. It serves as the portal to the AI Engine, AI Agents, and the Agentverse platform, enabling users to access a wide range of services and functionalities through a single interface.

Developers can leverage Fetch.ai’s AI Agents technology and encapsulate Large Language Models (LLMs), Machine Learning (ML) models, existing APIs, and other business logic to make services accessible via DeltaV. By developing and registering their AI Agents within the Agentverse: Services, developers can ensure that their services are retrievable and accessible through the DeltaV chat interface.

Collaboration with Bosch and Expansion into the Economy of Things

Fetch.ai and Bosch have expanded their cooperation by establishing the Fetch.ai Foundation to research, develop, and commercialize Web3 technologies. This collaboration aims to create machines that organize themselves in networks and make economic decisions, enabling objects in the Internet of Things (IoT) to engage in economic interactions, such as an electric car negotiating electricity prices.

AI and Machine Learning Capabilities

The integration of AI and machine learning into the Fetch.ai network is a pivotal feature of the platform. Fetch.ai employs sophisticated machine-learning algorithms and efficient data-handling mechanisms. These technologies empower Autonomous Economic Agents (AEAs) to learn from interactions, optimize decisions, and autonomously execute complex tasks. The network’s machine learning capabilities allow for the analysis and prediction of trends, enhancing the effectiveness of AEAs in various applications like supply chain management, energy distribution, and financial services.