What Andrej Karpathy and your legal team both get right about vibe coding
Both sides are solving the same problem from opposite ends, and they're missing each other in the middle.
When engineers define "AI-ready," legal inherits the error
Encoding contracts is a legal choice
The Governance Gap: What Harvey Misses
Autonomous agents are changing legal
Death by Dialogue: The Case for Killing the Legal Chatbot
Legal work needs density, not dialogue.
AI agents fail because they don't know what "customer" means at your company
AI agents fail because nobody defined what "customer" means in your business. Ontology infrastructure provides semantic guardrails that technical controls alone can't deliver.
Agentic AI doesn't need better models — it needs a data constitution
Agentic AI's real failure point isn't the model — it's the data pipeline. When agents act autonomously on corrupted data, output guardrails can't save you. Your data needs a constitution, not better prompts.
The vibe coding guy says vibe coding is passé
Karpathy says vibe coding is passé. The new term is "agentic engineering" — and for legal and product teams, the distinction is a governance question, not a branding one.
AI models that debate themselves are dramatically more accurate
Google research shows AI models that simulate internal debates dramatically outperform those that reason in monologue. For governance teams, the implication is clear: if dissent drives accuracy, hiding the chain-of-thought undermines trust.
