Friday, June 5, 2026
Agentic Daily
Stanford just delivered peer-grade evidence that AI writes better legal answers than law professors.
In a blind comparison of nearly 3,000 paired responses across 16 professors, Stanford Law School found AI-generated answers were rated higher 75% of the time and flagged as potentially harmful four times less often (3.5% vs 12.1%). Retire the "but AI hallucinates" line in your next vendor conversation — the data has changed, and your tutorial / training stack should change with it.
Tokens cannot be destroyed, only who pays for them changes.
Artificial Lawyer models law-firm AI spend as Harvey- and Legora-class tools move from learner-driver pilots to firm-wide deployment, and the projected line items dwarf today's enterprise-license fees. Get a tokens-per-matter unit-economic into next quarter's budget review before the question lands in your COO's office.
The week AI stopped being something you deploy and became something you can't account for
Anthropic admits Claude now writes most of its own code, bots just beat humans for control of the open web, and finance teams can't reconcile a token bill — the through-line is loss of human-legible accounting, not loss of control.