Testing Fable 5.1
Fable 5.1 jumped to 52.6% on an agentic science benchmark. I tested it by building a 3D brain in the browser. The model is capable. Law firms still cannot use it.
Vision
I work where legal services, AI, and transformation meet. I've built and scaled legal technology, helped lead two startups to exit, and now put AI to work at a global law firm.
Current Focus
Director, AI Applications at Pillsbury. I lead the firm's portfolio of AI applications, from evaluation and vendor strategy through rollout, adoption, and measurement.
Recent Thoughts
All essaysThe Sensation of Being Understood
LLMs did not solve understanding in the philosophical sense, but they solved the adoption problem by producing the sensation of being met in your own language.
The Industry AI Wasn't Supposed to Crack
Legal was supposed to be AI-proof. Slow procurement, conservative buyers, a profession built on human judgment. Now it's the #2 enterprise AI vertical at $500M — and what happened there tells you exactly what's coming for every other industry.
Experiments
AllCortex Atlas
A 3D replica of the human brain in the browser, built to test Claude Fable 5.1: real MRI atlas geometry, 130 researched cards, one file.
Prompt Craft Pro
Gamified prompt engineering platform that crossed 1,000+ active users during the Christmas launch period.
Fringe Legal
Seven-year legal innovation platform spanning podcast, community, and global events.
Published Work & Speaking
View allCharting AI Maturity
ILTA Peer to Peer •
Why Rejection Is Higher on Your Best Days
Taking LegalTech To Market •
AI Alone Isn't Enough: What Operators Are Seeing on the Ground
Fringe Legal Presents Bots @ Work •
SKILLS Annual Report 2026
SKILLS Summit •
A Practical Guide for Running Better Legal Tech Pilots
Singapore Law Gazette •
Walk Through: Lupl - Task Management + Workflow Automation
Artificial Lawyer •