150,347 validated knowledge triples per GPU-hour — the Understanding Efficiency Ratio reported in the framework paper , against the literature baselines it cites for fine-tuning and pre-training. Read the paper
This is the post where I'm supposed to be careful. The one where the lawyers would say "don't claim your AI is conscious" and the engineers would say "don't anthropomorphize
We have 123 research papers in various stages of completion. Seven are ready for journal submission. Nine more are in validation. The rest are in draft, design, or holding for
When someone asks us to build a NuSy being for a new domain — say, a clinical guideline specialist for diabetes management — the process looks nothing like training a
If you're building AI systems — agents, RAG pipelines, fine-tuned models, anything that learns — you already know that unit tests are necessary and completely insufficient. Tests tell you whether
Every NuSy being starts its life the same way. Not with training data. Not with weights. Not with a system prompt. With a story. "Once upon a time, in a
Here's a question nobody has a good answer to: How do you measure whether an AI system actually understands something? Not "can it generate a plausible answer." Not "does it
Two years ago, I ran a human development team. Good engineers. Agile ceremonies. Two-week sprints. We shipped maybe one or two features per sprint — sometimes three if the wind
Here's a question that haunts every organization: where does the knowledge go when people leave? It goes into Confluence pages nobody reads. SharePoint folders nobody can find. Slack threads that
I've been building AI systems since 2000. Clinical decision support at first — the kind where you wrote rules by hand and tested them against patient data and prayed they