The head of Meta's AI Safety & Alignment division watched OpenClaw nuke her whole inbox... скачать в хорошем качестве

The head of Meta's AI Safety & Alignment division watched OpenClaw nuke her whole inbox...

"I had to RUN to my Mac mini like I was defusing a bomb." - Summer Yue, head of safety and alignment at Meta Superintelligence. She asked OpenClaw to check her email inbox and see what can be cleaned. Told it specifically to "not action on anything until I approve." OpenClaw decided to nuke 200+ emails. By the time Summer realized what's happening, she kept telling OpenClaw to stop. OpenClaw did NOT listen and it only stopped when Summer literally unplugged her Mac Mini. Too many people are too focused on "what can the AI do?" that they missed the more important question: What have you done to PREVENT misaligned AI actions? I will take RELIABILITY over 'sexy' any day. ▶ If this helped you, hit LIKE and SUBSCRIBE. ───────────────────────── My Two Most Valuable Assets - Yours FREE: ❶ Join My FREE AI Community: https://augmented-intelligence.app.cl... ❷ Try NEO for FREE here: https://meetneo.app Get free credits in your account. ───────────────────────── 👋 About This Channel I create content showing what actually works in business and AI. Topics: AI processes, productivity systems, and implementation frameworks. Subscribe for more. 🔗 Connect With Me Twitter:   / ftayai   Facebook:   / felix.tay.3