Have you ever been to a film and as you walked out of that cinema you knew it had changed you in some profound way and that it would remain foremost in your mind for weeks to follow? Recently, I was fortunate enough to be asked to speak at DKGcon, in Amsterdam with my good friend Charles Ivie at Amazon Web Services (AWS) to represent the #DPROD initiative. I felt changed in a similar way.
At DKGcon, alongside 100 builders and 20,000 live-stream viewers, I learned about the movement that is creating a trusted Internet of data for artificial intelligence. You may be aware that many LLM models have been trained on inferior data scoured from across the public internet and consequently compound their proclivity for hallucination with poor data sources. To the extent that they have overcome this, mainly LLMs have resorted to manually curated data. But that alone cannot solve the challenges because statistical AI lacks true context and reasoning unless augmented by other models and technology like Graph RAG.
Branimir Rakic, Tomaz Levak, Žiga Drev, have created OriginTrail, a ground-breaking platform combining semantic web and blockchain to create a trusted, community driven economy for data. It is designed from the ground up to be compatible as a reference for LLMs and GenAI, yet is structured in a way that enables reasoning and inference based on knowledge graph (a.k.a. symbolic AI). The use of blockchain to underpin the network enables high levels of trust and provenance, empowering the community to collaborate to ensure quality and provide a powerful scalable economically viable platform to restore trust in data.
The conference focused on showcasing the great work from many of the community’s builders. These are truly inspirational individuals creating novel applications and networks of data assets (a.k.a. paranets) that enable trusted and truly innovative services.
I was very lucky to be able to share my knowledge of the Data product ontology DPROD alongside Charles Ivie and the supremely talented Marko Brkic who demonstrated it deployed in the DKG. DPROD was put together by many talented ontologists from around the world, coordinated by my dear friends, Tony Seale & Jacobus Geluk. Combining a metadata standard for data products with the DKG enables a far greater opportunity for interoperability straight out of the gate for any knowledge assets being created.
It was an honour to share my knowledge but, in truth, I came to learn and the conference did not disappoint. I found a truly inspirational community of people wanting to change the world for the better. I was astonished at the creativity, warmth and sense of community from all the people I was so lucky to have met through this event. I would wholeheartedly recommend finding out more. Too many names for LinkedIn to allow me to add here so names of my new and existing friends in the comments. #DKGcon2024 Ortecha, OriginTrail graph.build #dkg #knowledgegraph #datamanagement
