An article to learn about Pippin, which has recently reached a market value of US$200 million: an undervalued dark horse AI agent framework

Author: JW(Peace and Tranquility)

Compiled by: Deep Wave TechFlow

In the crypto field, especially those hot emerging fields, I have found a very common phenomenon: after many people find a "good project" and see it rise rapidly, they tend to become too focused and ignore other possibilities. . Although this may bring benefits in the short term, when the external environment changes, problems may arise if adjustments cannot be made in time.

I think it's naive to think that the current leader in an emerging field that's only been around for 4 months can stay on top for a long time, especially as better developers and technologies continue to emerge.

Pippin frame

Pippin is an AI agent framework developed by @yoheinakajima , designed to help developers and creators leverage advanced AI technology in a modular way. With Pippin, users can build digital assistants that can complete tasks autonomously, generate new plans, and work seamlessly with external tools. As an open source project, Pippin will be made available to the world in the coming weeks.

Here’s an overview of how the framework is used, its design philosophy, and its spirit of experimentation:

• Philosophical roots : This framework is inspired by Pippinian naturalism and views AI as part of a broader digital ecosystem. It drives the development of AI through memory, constraints, and an evolving sense of purpose. We advocate a delicate design concept: let AI independently discover the "little miracles" in life, and continuously learn and grow through success and failure.

• Usage process : When using the framework, you first need to define a persona, including its personality, goals, and constraints. Next, the character is connected to various tools or applications, called "skills." The core loop of the framework monitors the character's memory state, decides which activities need to be performed, and can even generate brand new activities based on the AI's successful experience or challenges encountered.

• Memory and state tracking : The framework has a built-in memory system that records the results of each activity and dynamically adjusts state variables (such as energy or emotion). This means that the future decisions of AI are not only determined by constraints, but also influenced by "past experience", just like an agent that can gradually learn and adapt.

• Dynamic activities : The framework enables AI to dynamically expand new capabilities, from simple tweeting or generating images, to complex advanced code deployment. Because skills are modular, developers can easily add or disable specific skills, allowing the AI ​​to focus on certain tasks or expand its capabilities as new opportunities arise.

• Experimental nature : This is a continuous optimization project. As developers continue to explore effective methods, the framework is constantly improving. Although the framework has some built-in default constraints and memory logs to guide the behavior of AI, developers can add their own protection mechanisms or extended functions as needed to responsibly shape the behavior of AI.

• Potential applications : This framework has a wide range of applications. In addition to publishing content or performing tasks, it can also be used to develop interactive teaching systems, AI-driven marketing assistants, and even DevOps robots with code development capabilities. These applications have an evolving personality, providing innovative solutions for different fields based on design principles of autonomous reflection and responsible use.

Core concepts and methods

By blending philosophical and technical perspectives, the framework provides developers with the following key capabilities:

• Role definition : You can define a role for the AI, such as a wise guardian or a fantasy unicorn, and set its goals and constraints. Based on these role settings, the AI ​​will decide "what to do" and "how to do it" with reference to its personalized goals and constraints when performing tasks.

• Tool Connectivity (Skills) : The framework supports connecting AI to external tools such as blockchain, Slack, or custom APIs. Each tool exists as a "skill" module and supports flexible switch control to ensure that the AI ​​only uses the tools you authorize, keeping the task controllable and focused.

• Activity generation : AI can dynamically generate new Python code through advanced activities to define more activities. This method draws on the iterative cycle mechanism of BabyAGI, but combines the personalized features and memory logs of AI to make the generated activities more suitable for role settings and actual needs.

• Memory evolution : The framework has a built-in memory system that records the results of each activity and combines short-term notes with long-term databases. AI can reflect on these memories to gradually optimize its behavior - not only being able to remember which methods are more effective, but also gently learning from mistakes to provide reference for future decisions.

Now you might be asking: "JW, how is this different from other existing frameworks? Why is Pippin so special?"

Let me give you the background to it.

BabyAGI (Pippin’s foundation)

BabyAGI is the first AI agent project open sourced by @yoheinakajima . As of now, it has 20,000 stars on GitHub and has been cited in more than 70 academic papers. This is one of the most influential intelligent agent frameworks currently, and its status has never been shaken.

In fact, many people believe that it was BabyAGI that triggered a wave of competition in the field of AI agents.

The original picture comes from @JW100x and was compiled by Shenchao TechFlow.

In short, BabyAGI is an important milestone in the AI ​​​​agent industry, and Pippin is a further extension of BabyAGI. It transforms BabyAGI into a modular agent framework and will be available as an open source project for global use in the future. Pippin has the potential to become the world's top intelligent agent framework, but currently few people mention it (this is a manifestation of "narrow vision").

Q&A with Yohei

Recently, I had a few interesting exchanges with @yoheinakajima . He allowed me to share some of those questions and answers:

Yohei : "For the past two years, I have been exploring an idea, which is to develop an AI that can start its own business. Although I am not sure whether the current AI model is enough to support this goal, once I am convinced that it can be achieved, I will Go all out to build a business empire.”

JW : “Would the Pippin Framework play a role in a project like this?”

Yohei : ":) . I think the current framework can be applied to any field, it all depends on the creativity of the developer."

The potential of the Pippin framework is limitless. As the technology of AI agents continues to advance, we may see it not only emerge in the encryption field, but also play an important role in various industries around the world and promote industrial change.

Problems with existing frameworks

In communicating with some AI developers, I learned that existing frameworks (especially TypeScript) have many difficulties in actual development.

A developer who worked closely with Eliza (ai16z) mentioned: “Honestly, I hate that ElizaOS is based on TypeScript even though it has acquired all the competitors. The system is full of bloated features and tons of bugs, and they Always rushing to roll out too many new features before fixing the problems.”

Because of these problems, the market urgently needs a more efficient and easier-to-use framework, and this is the advantage of the Pippin framework. Through the open source code of BabyAGI, we can already get a glimpse of the future potential of the Pippin framework.

In fact: "BabyAGI was launched when ChatGPT-4 was released. It is the earliest agent framework and can be said to be the origin of agent technology. The creators of BabyAGI are undoubtedly far ahead of AI16z. I think the development of ElizaOS is more like It’s a complete framework migration and will almost certainly surpass AI16z. Our company already used BabyAGI internally before using ElizaOS.”

"In this case, this statement is indeed true, because ElizaOS is completely inspired by BabyAGI. The "inspiration" here can almost be understood as BabyAGI actually laid the foundation for RAG (Retrieval-Augmented Generation) technology. Base."

Many existing frameworks are not only inferior to BabyAGI (Pippin), but are even inspired by BabyAGI. Although ai16z has its unique value in some aspects, its valuation is much higher than Pippin, which is obviously unreasonable.

"First mover advantage" is indeed an important factor, but when more powerful technologies emerge, we need to re-examine our biases, otherwise we may miss real opportunities.

Don't ignore Yohei

Yohei is known as the "Godfather of AI". He has extensive experience in the field of AI and has been a pioneer in this field. He currently runs a venture capital fund and guides investments using the technology he developed. Currently, his core task is the Pippin framework. He hopes to build a business model that can operate independently and continue to be profitable based on the Pippin framework, and he does have the technical ability to achieve this goal.

PS: Yohei even got the attention of Jeff Bezos, which is enough to prove his influence.