Войти
  • 102249Просмотров
  • 2 дня назадОпубликованоInternet of Bugs

AI May DOOM humans After All. I may have been wrong.

#AI #Cybersecurity #SoftwareEngineering #MoltBOOK #TechHype #AIAgents #Coding #MoltBOT #OpenClaw "The singularity is not happening. Fast take-off is not starting. But we might still be doomed—just not for the reasons you think." In today’s video, we’re diving into the "vibe-coded" chaos of MoltBOOK, the social network for AIs that has industry luminaries acting like they’ve seen the face of God. From world-readable databases to 48 security vulnerabilities in two weeks, I’m breaking down why this "social experiment" is actually a masterclass in how not to build software. We’ll debunk the "AI religion" hoax, look at the inherent insecurity of AI agents, and discuss why the tech community's uncritical response to this mess is the real danger. In this video: * What is MoltBOOK (and why the name keeps changing)? * The security nightmare of "vibe-coding." * Why industry experts are falling for obvious hoaxes. * A challenge for coders who want to actually understand how these agents work. Build your own Claude Code Challenge: // My Videos on AI Agents being Inherently Insecure 0:00 - The Singularity may be... Vibe-Coded? 1:33 - Why Humanity may (Actually) be Doomed 2:46 - What exactly is MoltBOT / OpenClaw? 4:10 - How AI Agents Work (And Why They Aren’t Safe) 5:49 - MoltBOOK: A Command and Control Infrastructure for Hackers 7:53 - The Hype Machine: Experts vs. Reality 8:21 - BREAKING: OpenAI’s Newest Hire ? 9:52 - Debunking the "AI Religion" and Doxxed Owner Hoaxes 12:04 - The Security Disaster: World-Readable Everything 15:15 - Why I Actually Dig Into the Code 16:14 - Challenge: Build Your Own AI Agent 18:58 - Final Warning: Let’s Be Careful Out There