ASIMOV at SEMANTiCS 2025

I've just got back from a week's trip to lovely Vienna for SEMANTiCS 2025 (aka the 21st International Conference on Semantic Systems), one of the premier annual knowledge graph research conferences in Europe. I delivered a talk about the ASIMOV platform, our polyglot development platform for trustworthy neurosymbolic AI that is now available in prerelease form—and promises over this next year to go on to superempower the work of many of the researchers in the audience.

I also presented a quick one-minute demo of our consumer-targeted ASIMOV personal intelligence app that we are building on top of the platform itself to help individuals, families, and teams deal with the cognitive overload of modern life, especially that inconvenient fact that each our social graphs now far overflow our wetware capacity as approximated by Dunbar's number:

My talk and demo was well received, with lots of audience interest and so many questions that I worried I was monopolizing too much of the time in the track! It was particularly gratifying to hear that our platform is "like a dream come true"—which is of course why we built it: we share that dream. Over the course of the conference I helped a number of people get set up with the ASIMOV platform on their laptop, contributing towards our initial user base as well as my understanding of user needs. ("Do things that don't scale!")

Beyond catching up with more than a dozen people and making a number of new friends, there were several pertinent research results presented at the conference that I've added to my research pile. From the presentations I attended myself, Professor Michael Färber's invited talk, "LLMs and Knowledge Graphs for Science: From Papers to Insights", was spot on in its relevancy to our work. It's certainly not only scientists who need trustworthy machine intelligence, but at the point when scientists could be satisfied with answer quality and factual fidelity, perhaps so could the rest of us—and as Prof. Färber showcased, the path to there goes through the neurosymbolic synthesis of language models augmented by knowledge graphs:

I'm looking forward to SEMANTiCS 2026 next year in Ghent, Belgium: a charming historic European city also hosting the prolific knowledge graph research group IDLab at Ghent University, some members of which I was able to catch up with in Vienna. It's definitely been too many years since I last visited them. Until then, also looking forward to following up on several research collaborations discussed in Vienna as well as catching up with many of the same fine folks at various upcoming annual research conferences next in Nara, Dayton, and Hersonissos, for starters!

Outside the conference itself, the last day of my trip to Vienna I also had the rare opportunity to catch up with Vlad Frolov, toolsmith extraordinaire and my former colleague at NEAR Protocol: