// MODEL OPTIMIZATION AND PROMPT SYNTAX TERM
Multi-Agent
An AI system where multiple independent AI models or "agents" work together, each with a specific role, to solve a complex problem.
TECHNICAL DEFINITION
A multi-agent AI system involves multiple autonomous large language models (LLMs) or AI entities, each with distinct roles and objectives, collaborating or interacting to achieve a common goal or solve complex problems through communication and task delegation.
BACKGROUND
Prompt engineering is the process of structuring natural language inputs to produce specified outputs from a generative artificial intelligence (GenAI) model. Context engineering is the related area of software engineering that focuses on the management of non-prompt and prompt contexts supplied to the GenAI model, such as system instructions, metadata, API tools and tokens.
READ MORE ON WIKIPEDIASYNONYMS & ALIASES
- Collaborative AI
- Agentic AI
- Distributed AI
USAGE NOTE
Multi-agent systems are being explored for complex tasks like scientific discovery or strategic planning.
DEVELOPERS
Organizations developing technology related to Multi-Agent.
Microsoft Research actively develops frameworks like AutoGen, which enables multi-agent conversations and prompt-based orchestration for various tasks, showcasing a strong focus on AI engineering and prompt design for agent collaboration.
DeepMind has been a pioneer in multi-agent reinforcement learning and designing complex agent systems that interact and learn from each other, often involving sophisticated engineering of their goals and communication protocols.
OpenAI explores the capabilities of large language models as agents and develops methods for orchestrating multiple agents to perform complex tasks, heavily relying on advanced prompt design and AI engineering principles for their interactions and decision-making.
Anthropic focuses on AI safety and alignment, including research into multi-agent systems to understand and control emergent behaviors. Their work involves careful prompt engineering to ensure agents interact safely and according to ethical guidelines.
Meta AI conducts extensive research in multi-agent reinforcement learning, social AI, and collaborative agents, developing systems where agents interact, communicate, and learn together to achieve common or individual goals.
Various research groups within CMU's School of Computer Science and Robotics Institute are leaders in multi-agent systems, human-agent interaction, and autonomous systems, developing foundational technologies for agent coordination and communication.
AI2 conducts fundamental and applied AI research, often including projects on multi-agent systems, knowledge representation, and reasoning that involve the engineering of agent architectures and their interactive behaviors.