Best of ai-agents — 2024

  1. 1
    Article
    Avatar of medium_jsMedium·1y

    AI Agents: How to build Digital Workers

    AI agents are transforming software development by integrating reasoning, decision-making, and action-taking capabilities. Leveraging large language models (LLMs), these agents can autonomously perform complex tasks, evolving from passive responders to intelligent systems. Key components include perception, reasoning (brain), memory, knowledge, and action through APIs. Designing an effective agent involves defining its role, tasks, memory, knowledge sources, and tools. When facing complex problems, a team of specialized agents working together can be more effective. Deployment considerations include security and operational tracking.

  2. 2
    Article
    Avatar of medium_jsMedium·2y

    Agentic AI: Creating An AI Agent Which Can Navigate The Internet

    WebVoyager is an innovative AI agent designed to navigate and interact with the internet using annotated screenshots and textual inputs. Leveraging large multimodal models (LMMs), it mimics human web browsing behavior by using visual cues, enabling it to perform tasks like clicking, typing, and scrolling autonomously. This agent showcases the advanced capabilities of GPT-4 models and underscores the significance of multimodal inputs for sophisticated web interactions.

  3. 3
    Article
    Avatar of baeldungBaeldung·1y

    Implementing an AI Assistant with Spring AI

    This tutorial delves into the features of Spring AI to create an AI assistant using LLMs like ChatGPT. It highlights the key functionalities, including context-aware response generation, structured output conversion, and integrating with Vector DBs. The process involves setting up necessary dependencies, creating relevant tables, and implementing callback functions. Common concerns like data privacy and maintaining conversational states are addressed using Advisors APIs. Examples demonstrate how to build a chatbot in a legacy Order Management System, showcasing practical applications of these concepts.

  4. 4
    Article
    Avatar of hnHacker News·2y

    mem0ai/mem0: The memory layer for Personalized AI

    Mem0 enhances AI assistants and agents with an intelligent memory layer, enabling personalized interactions by remembering user preferences, adapting to individual needs, and continuously improving over time. It uses a hybrid database approach to manage and retrieve long-term memories, which enhances the personalization and relevance of AI responses. Mem0 supports applications in customer support, healthcare, education, and more. It offers both a managed hosted solution and an open-source package for easy integration.

  5. 5
    Video
    Avatar of ibmtechnologyIBM Technology·2y

    How to Build a Multi-agent AI System

  6. 6
    Article
    Avatar of taiTowards AI·1y

    LLM Agents and Agentic Design Patterns

    Agentic AI is revolutionizing AI by enabling autonomous, distributed intelligent agents capable of real-time decision-making and dynamic problem-solving. A recent Berkeley course on LLM Agents offers deep insights into their history, reasoning patterns, and safety. Key highlights include the ReAct pattern for systematic exploration, importance of memory in LLMs, and the advantages of multi-agent systems for complex tasks. However, high costs and latency remain challenges. The field is rapidly evolving with new frameworks like Microsoft’s Autogen and Langchain's Langgraph.

  7. 7
    Article
    Avatar of freecodecampfreeCodeCamp·2y

    Create AI Assistants with OpenAI's Assistants API

    Learn how to create AI assistants with OpenAI's Assistants API. The course covers the basics of function calling, knowledge retrieval, and code interpretation. It includes hands-on projects and examples in Streamlit.

  8. 8
    Article
    Avatar of medium_jsMedium·2y

    OpenAI Swarm : A new Multi AI-Agent framework

    OpenAI has introduced Swarm, a lightweight Multi-Agent Orchestration framework designed for educational purposes. Swarm features agents with specific instructions and callable functions, operating statelessly unless context variables are explicitly used. Agents can hand off control to other agents, facilitating complex interactions with minimal functionalities. The post provides a basic example to demonstrate handoffs between agents.

  9. 9
    Article
    Avatar of ghblogGitHub Blog·2y

    What are AI agents and why do they matter?

    AI agents use large language models and additional algorithms to autonomously perform tasks, enhancing capabilities like software development and booking plane tickets. They can plan, use memory, and query APIs, making them versatile tools to minimize mundane tasks for developers. Ongoing challenges include predicting AI behavior and explaining outputs. Tools like GitHub Copilot Workspace are evolving to make developers more productive by enabling collaboration with multi-agent systems that handle complex tasks.

  10. 10
    Article
    Avatar of langchainLangChain·1y

    Command: a new tool for building multi-agent architectures in LangGraph

    Command is a new tool in langgraph designed to simplify the communication within multi-agent systems. It allows for the creation of edgeless graphs, enabling nodes to dynamically determine their subsequent nodes and states. This enhances the flexibility and control over multi-agent architectures, particularly in scenarios involving agent handoffs.

  11. 11
    Article
    Avatar of hnHacker News·2y

    AI multi-agents with corporate structures

    The performance of AI agents can be significantly impacted by organizing them based on corporate structures found in big tech companies. Competitive teams like those in Microsoft and Apple outperform centralized hierarchies seen in Google and Amazon. This experiment suggests that just as human organizational structures shape problem-solving approaches, so too can the structured interactions of AI agents. Although adding more agents improves accuracy to an extent, future advancements will likely depend on enhancing agents' logical reasoning capabilities and providing better tools.

  12. 12
    Article
    Avatar of tdsTowards Data Science·2y

    LLM Agents Demystified

    LightRAG provides an easy-to-implement solution for building autonomous agents capable of reasoning, planning, and acting. The ReAct Agent paradigm involves sequential steps of thought, action, and observation to solve user queries. The ReAct Agent class orchestrates a planner for generating responses and a ToolManager for managing tools, including LLM fallback options. Customization options are available, such as modifying templates and providing examples to ensure correct output format.

  13. 13
    Article
    Avatar of elixirstatusElixirStatus·1y

    agentjido/jido: A foundational framework for building autonomous, distributed agent systems in Elixir.

    Jido is a foundational framework for building autonomous, distributed agent systems in Elixir. It allows for the creation of smart, composable workflows that can adapt to their environment. Key features include composable actions, real-time sensors, adaptive learning, and built-in telemetry. It is designed for multi-node Elixir clusters and includes robust testing tools. Jido is currently under active development, with the stable API encompassing Actions, Workflows, Agents, and Sensors.

  14. 14
    Article
    Avatar of mlnewsMachine Learning News·2y

    MegaAgent: A Practical AI Framework Designed for Autonomous Cooperation in Large-Scale LLM Agent Systems

    MegaAgent is a new framework designed to enhance LLM-powered multi-agent systems by enabling dynamic task splitting and parallel execution without predefined Standard Operating Procedures. It features a hierarchical structure that allows tasks to be divided and managed by specialized agent groups, leading to significant improvements in scalability and efficiency. MegaAgent's architecture ensures that complex tasks can be completed with high accuracy by facilitating real-time communication and coordination among many agents. Experiments have demonstrated its superior performance and potential for various applications, including policy development and game creation.

  15. 15
    Video
    Avatar of TechWithTimTech With Tim·1y

    AI Agents Are Taking Over... Here's What You NEED to Know

    AI agents are advanced AI-driven systems that can autonomously complete tasks, make decisions, and adapt to new information. They are powered by large language models (LLMs) like GPT or Llama 3 and integrate technology like APIs, databases, and feedback loops to function effectively. Nvidia's RTX AIPCs support these AI agents for faster and better performance. The post discusses how to set up and use AI agents on your own computer, highlighting the 'Anything LLM' application that allows local experimentation with various LLMs and AI agent skills.

  16. 16
    Article
    Avatar of gopenaiGoPenAI·2y

    Orchestrating Intelligence: A Deep Dive into AI/ML Agents and Frameworks

    AI agents, which can operate sequentially, in parallel, or hierarchically, bring a new level of dynamism and complexity to the AI field. Various frameworks support these agents, including LangChain, OpenAI, Google Gemini, CrewAI, and AutoGen, each offering unique features for creating and deploying AI agents. The choice of framework and type of agent depends on the project’s specific requirements, whether it involves simple step-by-step tasks, handling multiple inputs simultaneously, or managing complex, multi-step processes.

  17. 17
    Article
    Avatar of medium_jsMedium·2y

    DSPy vs Conva.AI : Building the Best AI Assistant

    The post compares DSPy and Conva.AI, two AI assistant building platforms, by using them to build assistants for online shopping, payment management, and venue bookings. Conva.AI, with its user-friendly interface and automatic context extraction from URLs, proves to be more consistent and easier to integrate. DSPy offers more control and customization for users with programming expertise but shows inconsistency in outputs. For users without coding skills, Conva.AI is recommended, while DSPy suits those with coding proficiency seeking advanced customization.

  18. 18
    Article
    Avatar of gettingstartedaiGetting started with AI·2y

    Installing and exploring AutoGen Studio 2.0

    AutoGen Studio 2.0 is a web application that simplifies setting up, managing, and testing multi-agent AI workflows. It allows developers to easily create and manage AI agents, skills, models, and workflows. AutoGen is an open-source framework that enables the building of AI agents that communicate and collaborate to accomplish tasks. The possibilities for applications of AutoGen are limitless, including travel planning, reservations, customer support, and research and analysis. With AutoGen Studio, developers can quickly dive in without a steep learning curve. The AutoGen Studio UI provides sections for managing skills, models, agents, and workflows. Additionally, the AutoGen Studio Playground allows users to create new sessions and watch their agents communicate and behave to complete requests.

  19. 19
    Article
    Avatar of mlnewsMachine Learning News·2y

    Guided Reasoning: A New Approach to Improving Multi-Agent System Intelligence

    Guided Reasoning is a technique introduced by Logikon AI, where a guide agent assists client agents in reasoning through problems using a methodical approach. The guide structures the process by setting rules, posing questions, and evaluating responses to ensure accurate and explainable AI outputs. This technique divides cognitive work effectively, aiming for better problem-solving in multi-agent systems.

  20. 20
    Article
    Avatar of mlnewsMachine Learning News·2y

    Scalable Multi-Agent Reinforcement Learning Framework for Efficient Decision-Making in Large-Scale Systems

    Researchers from Peking University and King’s College London developed a decentralized policy optimization framework for multi-agent systems, improving scalability and decision-making efficiency in large-scale AI systems by reducing communication and system complexity. The framework uses model learning to enhance policy optimization with limited data and employs localized models for accurate state and reward predictions. Tested in diverse scenarios like transportation and power systems, it demonstrated superior performance, significantly reducing communication costs while improving convergence and sample efficiency. This scalable MARL framework shows potential for applications in advanced traffic, energy, and pandemic management.

  21. 21
    Article
    Avatar of substackSubstack·2y

    The "Egg Theory" of AI Agents

    The 'Egg Theory' of AI Agents explores the idea that AI products shouldn't completely remove humans from the loop and should instead include moments of human involvement or the illusion of control.

  22. 22
    Video
    Avatar of twoninutepapersTwo Minute Papers·2y

    Meta’s Llama3 AI: ChatGPT Intelligence… For Free!

    Meta has released Llama3, an AI chatbot assistant similar to GPT-4, which is performing well on coding tasks. Llama3 also has impressive performance on the tough science test GPQA. There will be a 400-ish billion parameter model available in the future. Llama3 is comparable to earlier versions of GPT-4 and is available for free. Google DeepMind's Gemini 1.5 Pro is also performing well and has a large context window. Overall, AI assistants are improving rapidly and open source models are back in the game.

  23. 23
    Article
    Avatar of hnHacker News·2y

    Let's Build AI

    Let's Build AI provides developer tools, model development, and autonomous agents for running AI models on your own data. It also offers workflow automation and cloud providers for image generation. Contributions are welcome on GitHub.

  24. 24
    Article
    Avatar of medium_jsMedium·2y

    From Basics to Advanced: Exploring LangGraph

    LangGraph is a new module within the LangChain ecosystem designed to handle complex workflows involving cyclical graphs and human-in-the-loop interactions. It offers extensive customization and control, making it suitable for advanced LLM applications, unlike the more high-level CrewAI framework. The post explores LangGraph’s capabilities by demonstrating how to build single-agent and multi-agent workflows, including error correction and user feedback integration. It contrasts LangGraph's flexibility with CrewAI's ease of use and prebuilt features, highlighting LangGraph's suitability for more advanced applications requiring detailed customization.

  25. 25
    Article
    Avatar of collectionsCollections·2y

    Llama-Agents: An Open-Source Framework for Multi-Agent AI Systems

    Llama-Agents is an open-source AI framework that simplifies the creation and management of multi-agent systems using an async-first approach and large language models. It features a distributed architecture, centralized control plane, standardized communication protocols, and supports human-in-the-loop processes. The framework is easy to deploy and scalable, offering built-in tools for monitoring and interaction.

See all ai-agents archives