Theoriq’s first Educational AMA Series has gone off with a bang! Hundreds of the Theoriq community tuned into hear our Head of AI Education Shingai Manjengwa and her team, Senior AI Solutions Engineer, Pourya Vakilipourtakalou and Alexander Marx, discuss Theoriq’s AI Agent Base Layer, as well as a look into the recently released Litepaper, and a deeper dive into what we mean by AI agents.

There was so much valuable information we had to share it with you all. Below you will find some important terms for our community to understand over the next few years, with a focus on AI Agent collectives, multi-agent systems and a use case of how AI Agents can already impact web3.

Let's get into it.

The most important term to understand first, is what we mean when we say an AI Agent?

but much different to your everyday chatbot assistant – this one can dynamically read documents, it can process vast amounts of information online, it can write and execute code, it can even make calculated decisions all by itself!

This is an AI Agent.

An Agent is a form of software that leverages the capabilities of AI to perform a specific function. These Agents use advanced AI models – such as Large Language Models (LLMs) – to perform specific tasks in which users can interact with them using natural language. Agents are able to access real-time data, interact with other software programs, and automate actions, like creating and sending emails or discovering and trading specific stocks.

More formally, at Theoriq, we define agents as follows:

“AI agents are autonomous software systems that leverage modern generative AI models to plan, access data, use tools, make decisions, and interact with the real world to perform specific functions.” Theoriq Team. (2024).  Theoriq: The AI Agent Base Layer.

Before diving into how agents work and collaborate, it's important to understand what powers them. Foundational AI models, such as LLMs, are advanced AI systems that are trained on vast amounts of data. They learn to understand and generate human-like responses by looking at many examples of data, which range from digital blogs, social media posts, Wikipedia pages, and many other sources. This type of analysis is called ‘training’.

From these examples, AI models, like LLMs, have developed the ability of prediction, where they can predict the next word in a sentence, or paragraph based on patterns learned from their ‘training’ data. This form of training, combined with Instruction Tuning and Reinforced Learning From Human Feedback (RLFH), have evolved LLMs from basic inhuman outputs, to the ability to write essays, answer questions, and even hold conversations. Users can now interact with LLMs through chat interfaces by entering questions or instructions to LLMs, which are called ‘prompts’. We have seen this happen with the growth of interfaces like ChatGPT.

So do these models know all human knowledge? No, their knowledge is limited to what they have learnt during their training. Which means they may not always be up to date with the latest information, and is why agents need access to tools like the internet to solve more complex tasks accurately and efficiently.

Now, imagine if instead of relying on one assistant, you had a whole team of specialists. This is where Agent Collectives come into play. In Theoriq, Agent Collectives are groups of AI Agents that collaborate to tackle more complex tasks.

Each agent in the Collective has a specific role. For example, one Agent might be great at analyzing data, while another is skilled at interpreting social media trends. Together, they can handle more complex tasks that would be too challenging for a single Agent.

Let's have a look at a specific example of how AI Agent Collectives could work in reality.

Let's say you want to analyze the latest trends in different Decentralized Financial (DeFi) markets. This is a laborious task that requires many hours of in depth research. But instead of doing this yourself, you can utilize  a group of AI Agents to create a Web3 Market Analysis Collective. This team of specialized AI Agents would include:

• A Router that delegates tasks to different agents depending on the task.

• A Social Media Specialist that analyzes social media data to learn market insights.

• A News Analyst that gathers and summarizes relevant news about the market.

• A Data Analyst that examines the heavy data analysis involved in trading data, Market Caps (MC), Total Value Locked (TVL), and liquidity trends.

Each part of the task is handled by the Agent best suited for it, making the whole Collective more efficient and supercharging your market analysis process. This Collective would boost your productivity, and free you up from examining an overwhelming amount of data. You could pick and synthesize the most useful information relative to your specific needs.

Using Agent Collectives has several advantages for users including:

• Efficiency and Specialization: By dividing tasks among specialized Agents, you can gather results faster and with more accuracy, instead of using primitive models that are trained only to perform one task.

• Collaboration and Creativity: Similar to human teams, AI Agents who work together can offer a more diverse range of perspectives and approaches, this enhances creativity and their ability to solve complex problems.

• Scalability and Flexibility: Collectives are inherently more scalable and adaptable than a single Agent. They can reconfigure themselves in response to new tasks, integrating new specialized agents as needed. This modularity allows for continuous improvement without having to overhaul an entire system.

• Interoperability and Composability: Theoriq's protocol ensures that Agents can seamlessly communicate and collaborate, regardless of their underlying implementations. This interoperability (Agents who can work together) allows for the dynamic discovery and composition of Agent Collectives, enabling more flexibility and more powerful AI agentic systems.

Theoriq is revolutionizing how AI Agents work together by creating a framework where they can form Collectives to tackle complex tasks more efficiently. By leveraging the strengths of individual specialized agents and empowering more seamless collaboration, Theoriq is paving the way for more powerful, adaptable, and responsible AI systems.

The governance of Theoriq's AI collectives is anchored in a decentralized, community-driven model that promotes transparency and accountability. By leveraging mechanisms like Proof of Contribution and Proof of Collaboration, the system ensures that the most effective agents are rewarded and aligned with community values. Stay tuned, as we'll dive deeper into this in the next blog of the educational series.

Whether you're new to AI or looking to deepen your understanding, exploring the world of AI Agents and their Collectives offers a fascinating glimpse into the future of technology. Dive into our X Spaces and Discord sessions to dive into the world of AI Agents and more.

Stay curious, keep learning, and together, let's unlock the full potential of AI!

If you want to have a deeper dive into all of the above, head over to our litepaper and read how we are building the future of AI Agent Collectives.

About Theoriq

Theoriq is a decentralized protocol for governing multi-agent systems built by integrating AI with blockchain technology. The platform supports a flexible and modular base layer that powers an ecosystem of dynamic AI Agent collectives that are interoperable, composable and decentralized.

By harnessing the decentralized nature of web3, Theoriq is unlocking the potential of collective AI by empowering communities, developers, researchers, and AI enthusiasts to actively shape the future of decentralized AI.

Theoriq has raised over $10.4M and is backed by Hack VC, Foresight Ventures, Inception Capital, HTX Ventures and more, and have joined start-up programs with Google Cloud and NVIDIA.

Website | X | Discord | LinkedIn