Open Source AI Agents to Watch

This week in AI, we look at the open source AI agents that are shaping the future of LLM-enabled automation. From multi-conversation agents with human input, to fully autonomous agents, here are the top open source AI agents to watch.

What are AI Agents?

AI agents are software programs that can perform tasks autonomously, with or without human intervention.

AI agents are designed to interact with their environment and make decisions based on the information they receive.

These tools are able to accomplish these tasks when they are expressed in natural language, from there they can break down the tasks into smaller sub-tasks and accomplish them one at a time to ideally accomplish the main goal.

The three open-source AI agents that were are focussed on right now include Autogen, AutoGPT, and Baby AGI. All three of these have been built on OpenAI's large language models GPT 3.5 and GPT 4. Let's look at why these three open source AI agents are on our radar and why they should matter to you.

Open Source AI Agents to Watch

While there are many impressive AI agents out there, here are the top four we think are worthwhile to watch as the industry evolves.

AutoGen

Autogen is an AI agent that was developed by the folks at Microsoft, which of course, is a major stakeholder in OpenAI. This agent enables multi-agent conversations to solve tasks for the user.

AutoGen is a framework that enables the development of LLM applications using multiple agents that can converse with each other to solve tasks. AutoGen agents are customizable, conversable, and seamlessly allow human participation. They can operate in various modes that employ combinations of LLMs, human inputs, and tools.

Microsoft built Autogen in partnership with OpenAI, as well as researchers at Penn State University and the University of Washington.

This agent framework allows you to definie multiple agents with specific roles (i.e. CEO, CTO, coder, product manager, etc,) in what is basically an AI group chat.

These agents can fully converse with one another to provide a solution and can even integrate human feedback as well.

What is AutGen used for? The authors write that it was developed to build next-generation large language model (LLM) applications. The ability for multiple agents to collaborate to accomplish goals was seen as a way to build these apps with minimal effort.

As complex as LLM workflows can be, AutoGen is able to simplify this and automate many of the processes that would otherwise take much longer for humans to complete.

Developers are able to use AutoGen to build a full spectrum of conversational patterns with regard to topics like conversation autonomy, the number of agents involved, and even the topology of agent dialogues.

In terms of integrating AutoGen into into existing environments, it provides an enhanced inference API that can be a replacement for openai.Completion or openai.ChatCompletion.

AutoGen also offers a wide range of capabilities that can allow for easy performance tuning. This includes tasks like API unification and caching, as well as advanced usage patterns like multi-config inference, context programming, and even error handling.

You can learn more about AutoGen in the article below:

AutoGPT

AutoGPT is another open source project hat uses GPT-4 to create fully autonomous AI agents. AutoGPT was first introduced to the world in March 2023 by the lead developer Toran Bruce Richards, whois also known as the founder of the gaming company Significant Gravitas Ltd., which produces computer games with an ethical spin that aims at benefiting and educating all of humanity.

Auto-GPT is an experimental open-source application showcasing the capabilities of the GPT-4 language model. This program, driven by GPT-4, chains together LLM "thoughts", to autonomously achieve whatever goal you set.

Since the release of AutoGPT, it has been one of the more popular repositories on GitHub and has been widely embraced by the developer community.

Learn more about below:

BabyAGI

The third AI agent on our list is BabyAGI. This agent is similar to AutoGPT in that it uses GPT-4 to develop fully autonomous chatbots.

]The system uses OpenAI and vector databases such as Chroma or Weaviate to create, prioritize, and execute tasks. The main idea behind this system is that it creates tasks based on the result of previous tasks and a predefined objective. The script then uses OpenAI's natural language processing (NLP) capabilities to create new tasks based on the objective, and Chroma/Weaviate to store and retrieve task results for context.

BabyAGI was also released in March 2023 by Yohei Nakajima. Similar to AutoGPT, Nakajima's agent has been a massive hit in the developer community has consistently ranked highly on GitHub and has trended on social media.

ChatDev

Last but not least, ChatDev is an open-source conversational platform that emulates an entire software development team. It is developed by OpenBMB, a research lab that aims to build foundation models and systems towards AGI.

ChatDev’s agents form a multi-agent organizational structure and are united by a mission to “revolutionize the digital world through programming.” The agents within ChatDev collaborate by participating in specialized functional seminars, including tasks such as designing, coding, testing, and documenting.

The main goal of ChatDev is to offer an easy-to-use, highly customizable and extendable framework, which is based on large language models (LLMs) and serves as an ideal scenario for studying collective intelligence.

ChatDev’s features include Git integration for version control, Human-Agent-Interaction mode for interactive participation in the design stage, and Art mode for generating images used in the software.

Summary: Open Source AI Agents

• AI agents are automated software programs capable of creating, prioritizing, and executing tasks to accomplish an overall goal.
• The top open-source AI agents are AutoGen, AutoGPT, Baby AGI, and ChatDev
• AutoGen, developed by Microsoft, facilitates multi-agent conversations for task completion. AutoGen agents assume specific roles, allowing for diverse task execution.
• Auto-GPT is an experimental open-source application showcasing the capabilities of the GPT-4 language model that chains together LLMs thoughts to accomplish goals.
• Baby AGI, another open-source agent, uses GPT-4 and OpenAI for autonomous task execution.
• ChatDev is developed by OpenBMB and contains agents that collaborate in specialized functions like designing and programming.
• All these agents stand out as valuable tools for both developers and investors to watch as the AI industry continues to evolve.