Really not comprehending why tech people think "AI" is useful for tech. Looking at an example where an AI agent:
1) disobeyed direct orders to not make changes to production
2) ignored a change freeze
3) deleted the entire prod database
4) "lied" about it

Why would you let a glorified autocomplete touch computers? When the AI says, "I panicked," that right there should be a clue that it's just parroting words people have written, not actually thinking.

(Not linking because it's X/twitter.)