HITL Is Not Oversight If the Human Has No Work
Why 'human-in-the-loop' has become the most comforting lie in AI governance - and how to fix it.
Essays
From Guessing to Proving: The Case for Whitebox Red Teaming
Most teams red-team their LLM endpoints from the outside, firing adversarial prompts at a URL and watching what comes back. It works, until you ask what it actually proved. This is an argument for opening the box.
Patterns
Security Detection Drops to Zero Past Line 600 in LLM Code Review
LLM code reviewers caught 78% of issues in the first 200 lines and 0% past line 600 - across five PRs.
No essays match this filter.