Notes — Patrick Colm Audley

PurRDF 1.0.0 is out.

LLM's generate one token at a time, but we usually consume them as turns - that's missing most of the fun!

I've started a new ontology project: GMEOW — the Global Metadata and Entity Ontology for the Web.

The recurring problem with useful personal agents is that the context they need already exists, but not in the tidy place a demo would like it to be.

As AI coding agents transition from passive autocomplete tools to autonomous contributors executing entire feature branches, we are racing toward a massive security blind spot: How do we prove a real human actually reviewed and verified agent-generated co…

The thing I keep running into with multi-agent setups is that the engineering principles a team actually cares about — how to handle errors, when to wrap shell calls, what counts as a critical path — live in a wiki page or a slide deck nobody reads.

Standard Graph Neural Networks are structurally constrained when mapping complex text attribution: linear aggregation in flat Euclidean space inevitably forces semantic drift.