Janet Adams presented an overview of SingularityNET’s emerging Knowledge Layer strategy at Rare Evo 23 in Denver. Today, we are excited to share more details.

Also available in Arabic, Spanish, Portuguese, French, Turkish, Indonesian, Chinese, Russian, Italian

Dear Singularitarians,

The next pivotal phase of growth of the SingularityNET platform and ecosystem is here — broadening SingularityNET’s scope to become the Knowledge Layer of the Internet.

This stage in SingularityNET’s growth goes beyond a strategic initiative; it’s a vision that places the platform at the forefront of the ongoing quest to decentralize AI, and emerging decentralized AGI, for broad human benefit.

While the benefits of LLMs and generative AI are increasingly exciting, their limitations are also becoming increasingly apparent. The convergence of structured knowledge with neural networks is emerging as the key to overcoming their shortcomings.

Uniquely positioned at the heart of this convergence, SingularityNET will prioritize the integration of Knowledge Graphs (KGs) and comprehensive Neural Symbolic tools for LLMs, with the goal of empowering developers to create the next generation of AI systems on the SingularityNET Platform.

SingularityNET is strategically positioned to become the Knowledge Layer of the Internet, leading the evolution of the emerging AI landscape while also laying the foundations for the Artificial General Intelligence (AGI) of tomorrow.

At the Cardano community event, Rare Evo, our COO Janet Adams painted the picture of this new Knowledge Layer, and the opportunities it presents. The blockchain Layer 3, the Internet of Knowledge that SingularityNET is architecting, will facilitate the evolution from the current data-driven AI to the future of knowledge-driven AGI.

Watch the full presentation on YouTube.

In this update, we’ll delve deeper into what this transition means for SingularityNET, the AI industry, and the future of humanity.

The SingularityNET as a Blockchain Layer 3 — The Knowledge Layer of the Internet

LLMs and other deep neural networks have revolutionized how people all over the world work, learn, and thrive; their ability to understand our needs and amplify our skills and imagination has made them the most quickly embraced technology in history, and with good reason. They offer us new possibilities in empowering people to achieve excellence via human-technology collaboration.

But these tools do have their limitations.

They’re not trustworthy for critical applications or ethical decision-making — they hallucinate, fabricate, and propagate bias. They lack transparency, they’re data and resource intensive, and they’re controlled by centralized entities with their own agendas.

These are limitations that cannot be overlooked. They prevent us from using these innovative tools effectively in our work and daily lives, limiting their own potential.

But there is a path to overcoming these challenges — by neural symbolic AI techniques, such as integrating Knowledge Graphs (KGs) into advanced LLMS and other deep neural networks.

Knowledge Graphs (KGs) are a form of structured data, which embody context and relational understanding within a dataset, resulting in more resilient data infrastructures that are grounded in reality.

Think of KGs as a family tree. Each person (node) on the tree represents a piece of information. The lines connecting them (edges) show how they’re related. Just like you can see who’s a sibling or a parent in the tree, in a knowledge graph, you can see both the indivdual units of information (the nodes), and how those units are related to each other (the links).

Neural symbolic AI brings together the best of both neural networks, like LLMs, and symbolic reasoning systems, like knowledge graphs, into one unified framework.

Decades of research and development fortify our pivot towards becoming a Layer 3 blockchain for knowledge — SingularityNET is the world’s leading authority in the integration of graphs, KGs, metagraphs, and the AtomSpace developed as part of the OpenCog project.

By building a thriving community of KG developers, we are setting the stage for the next phase of AI. This community will be central to building the foundations of the Layer 3 of blockchain, the knowledge layer of the Internet. Its contributions will be integral to the next steps of building fully neural symbolic LLMs, as well as building a critical foundation for the Distributed Atomspace, the hypergraph knowledge store at the heart of OpenCog Hyperon.

This focus on neural symbolic techniques, knowledge graphs, and knowledge graph tooling not only solves today’s AI challenges but also positions us to lead the evolution towards decentralized, benevolent AI & AGI systems.

Ultimately, these factors position SingularityNET to become the Knowledge Layer of the Internet — a blockchain Layer 3. Layer 1s and Layer 2s refer to the foundational blockchain protocols and scalability solutions, respectively, while a Layer 3 blockchain could comprise the information layer, adding contextual data and complex logic on top of the two base layers.

The Next Generation of AI Advances Is Here

As we move beyond the age of big data to an era of “structured data,” the role of knowledge graphs in enhancing AI systems becomes critical. It’s no longer about having raw data; the next phase is about having organized, relationship-driven, contextual information that can fuel more intelligent algorithms.

We are setting the stage for the next generation of AI systems to be built on knowledge; rather than just data.

Here are the three fundamental pillars the AI ecosystem of the future will be built upon:

Knowledge Graphs: Specialized KGs that augment LLM and AI capability will form a public infrastructure of structured knowledge — commonsense knowledge, genetic ontologies, regulatory considerations are just a small example of possible specialized knowledge graphs that developers will increasingly need.

Knowledge Graph tooling: A comprehensive set of Knowledge Graph (KG) tools for building, editing, and visualizing complex data structures will enable anyone to create, customize or understand KGs.

Third-party LLM connectors: To create an expansive, inclusive ecosystem, we will also introduce connector tools to facilitate the integration of third-party LLMs. These connectors make it easier for specialized models to be added to the platform, allowing for more fine-tuned and specialized AI capabilities.

Through its emphasis on data interaction, formal verification, and our aligned values of openness and decentralization, the Cardano blockchain will enhance our transition to the Knowledge Layer of the Internet and provide an architecture that will empower our community of AI developers.

These three pillars — Knowledge Graphs, Knowledge Graph Tools, and third-party LLM connectors — will set the groundwork for the greater vision of the emergence of AGI.

Creating Decentralized, Beneficial AGI — One Knowledge Graph at a Time

Imagine a world where the emerging power of LLMs is paired with tools that make them trustworthy, reliable, and grounded in reality.

Neural symbolic elements are the key to this future, and they will power a massive phase of growth for the SingularityNET Platform and marketplace.

The integration of neural symbolic tooling, and the development of our own Neural Symbolic LLM, Zarqa, will make SingularityNET the home of the most advanced AI in the world.

The neural symbolic architecture and knowledge-rich systems will offer an environment conducive for the growth of OpenCog Hyperon, and it will lay down the principles, technology, and community empowerment that AGI requires.

This new strategic roadmap, the new innovative knowledge graph tooling, and our community of dedicated developers, all come together to achieve our overarching objective — foster the growth of the SingularityNET platform, and ultimately achieving beneficial decentralized AGI using OpenCog Hyperon.

The Future is Already Here — Explore the Awarded Projects in Deep Funding Round 3

Deep Funding Round 3 has been an incredible journey for the entire community, and some of those proposals are already showing us the potential of leveraging Knowledge Graphs for building the next generation of AI tools.

Each phase of the Deep Funding Round 3 journey has played a pivotal role, ensuring the most promising projects, those aligned with the vision of SingularityNET and the community-approved guidelines of Deep Funding, received the support and funding they need to propel their growth forward.

You can learn more about the 43 awarded projects on the Deep Funding blog, and learn how they’re helping pave the way for the next generation of AI innovations in the SingularityNET ecosystem.

Deep Funding Round 3 Results