293 essays on what AI actually does to legal practice. Written while it happened, not after.
Everyone became an expert in the last three years.
I started writing days after GPT 4 shipped. Three years and 293 issues later, the field is loud and the slop is deep. This archive is the difference: timestamped thinking, tested claims, positions taken before they were safe to take.
New to Brainyacts, or sent here by a colleague?
The special briefing on the economics your pricing committee is avoiding. The thesis behind everything else here.
What that job posting really signals. Then the follow up, where I built the role's toolkit for a few hundred dollars.
The power dynamic between firms and clients is inverting. What partners should do while they still can.
Not by date. Six threads run through three years of writing. Each one reads in order.
The legal AI terms that matter, defined plainly, updated as the field moves. Link it. Forward it. Hand it to your associates.
Josh Kubicki. Professor, advisor, builder.
I teach at Indiana University's Maurer School of Law. I advise managing partners, chairs, executive committees, and practice group leaders on strategy: business models, practice group economics, competitive position, and what AI changes about all three. The essays here are the working notes behind that practice. Reporters, firm leaders, and GCs read them. Some even subscribe.