'I Used AI. It Worked. I Hated It' — AI Security Expert's Three-Week Claude Code Reckoning Gets 95 Comments

Hacker News·medium signal

M.T. Taggart, founder of The Taggart Institute and self-described 'anti-genAI' security expert, used Claude Code with Sonnet 4.6 to build a production certificate generation system in three weeks instead of six months. Despite the 3x speed gain and Claude catching security vulnerabilities he would have missed, Taggart describes the process as 'miserable' — development became reviewing and accepting changes rather than creating. He warns that 'human in the loop' processes encourage humans to exit the loop, and calls AI tools 'addictive substances' that alter cognition. The 40pts/95cmts ratio (2.375) signals intense practitioner debate.