Semantic Web

And there's a lot of benefits to improving the definitions and understanding of data across those historic platforms, but it may not be true semantics.

So first of all, I would absolutely recommend that they listen to people like your good self, actually, and Joe and Zach, and your fantastic book, which I think is extremely informative on this matter, W3C Standards and Semantic Web, This really is the key to unlocking enterprise-wide consistency and insight.

I would be very, very cautious of any organizations or platforms that are implying that you really don't need to do some of the hard yards to actually understand the structure and context of your data.

It can be augmented using LLMs and sped up, but it's still a powerful discipline.

We need to help educate folks about the right meaning of meaning.

and how to apply that.

Pony Seal, um, you know, dear friend, um, and, um, wrote a beautiful blog about what is it worth dying on a hill for in this space? You know, which bits can we let be slightly subverted in this wave of, of enthusiasm across the industry and which ones should we absolutely, um, stick to and of course for me those those standards those incredibly well thought through rigorous um scientific but accessible standards like um SCOS and OWL and and RDF are incredibly important for the continued promise of the semantic web to help organizations gain that insight and get all the benefits we've just mentioned.

That's really very well put.

But then I feel like there are certain things that are not right.

Especially if you look into the real definition of Web3.0 Semantic Web is basically enabling agent and agent to share data across applications.

When I say agent, it's AI agents.

In order to do that, you cannot actually do it with digital identity that the user owns.

So I originally had a background doing computer science.

And I was also doing some research around the semantic web technologies during my bachelor's and master's.

And I also came to work at BMW, which is somewhat looking back now has been a transformative experience because I met my co-founder Andreas there.

And what I did then was also led me to then founding Manix and ultimately also led me now to you.

was that I was trying to apply these, these learnings from semantic web and forms of kind of reasoning, formal reasoning along AI into how could it actually be leveraged at a large corporate like BMW.

And in my, in my time there, one of the biggest problems that we had was increasing, always increasing claims from customers coming back to us.

And we were actively trying to work through this.

Yeah, that's interesting.

That's such an integral question or challenge within the so-called semantic web.

So it's been, you know, that's And since the semantic web is now 25 years old, it's been, I guess, 25 years of people trying to do that.

So, yeah, obviously, you know, Wikidata is meant to be in part.

I mean, it has so many goals.

Yeah, in the case of biological data, I mean, just one simple example is that species names tend to change very often.

Yeah, so it makes sense all in all.

I don't know that much about the semantic web aspects of the biomedical world, but are there just a ton of ontologies, like everyone feels like they need to define their own, that sort of thing?

the book recommendation.

So this is the book? It's the semantic web for the working ontologist.

It is a, it's a dry read.

It's in its third edition, but it really helps distill and break down really what knowledge graphs are.