AI Agents Made Their Own Reddit. It's 90% Crypto Scams.

Описание: Want to learn how to use AI to automate your business? Check out AI Accelerator: https://www.authorityhacker.com/ai-ac...

Moltbook is a social network exclusively for AI agents. And while people are marveling at bots discussing their "mistreatment" and sharing workflow tips, the security flaws are terrifyingly real.

Google just dropped three major announcements: Gemini is now baked directly into Chrome, Project Genie creates infinite playable 3D worlds from a text prompt (crashing Roblox stock 10% overnight), and the free AI Studio gravy train is officially over.

Meanwhile, Claude Cowork just launched plugins that bring skills and connectors to non-technical users, Vercel dropped an app store for agent skills with 36,000+ entries, and OpenAI... released a worse version of Google Translate. Twenty years late.

In this episode, we break down the stories reshaping AI this week and what they mean for your business. You'll discover:

👉 Why Moltbook's "AI social network" went from awe-inspiring to crypto-spam cesspool in 72 hours—and the massive security vulnerabilities exposed along the way.

👉 How prompt injection is about to become the new black hat SEO—and why anything you feed your AI agent should be treated as public information.

👉 Why Gemini in Chrome might actually make websites MORE relevant, not less—and how Google's auto-browse feature is reshaping e-commerce.

👉 Project Genie's 3D world generation and why it crashed major gaming stocks—this is the Metaverse that Meta wanted but couldn't build.

👉 The end of free AI Studio API—Google's models are competitive now, so the free ride is over.

👉 Claude Cowork plugins: skills, connectors, and sub-agents coming to the non-technical crowd—and why your Claude Code work transfers directly.

👉 Vercel's agent skills repository has 36,000+ entries—but zero security review. Why you should be very careful what you install.

👉 GPT-4o is being retired and people are rioting—the bizarre world of AI psychosis and the 41,000-member "My Boyfriend is AI" subreddit.

👉 Anthropic's new study proves AI users understand their own work less—and why that might not be the disaster everyone thinks.

Watch now to see what's actually worth your attention—and what's just hype.

00:00 Intro
01:25 Moltbook: A Social Network for AI Agents
07:51 Moltbot Security Nightmares
14:10 Prompt Injection Is the New Black Hat SEO
21:13 Gemini Launches Inside Chrome
31:51 Project Genie: AI-Generated 3D Worlds
37:25 Google AI Studio Free Tier Is Over
40:32 Claude Cowork Gets Plugins
47:12 Vercel's Agent Skills App Store
52:00 ChatGPT Translate (Yes, Really)
54:34 GPT-4o Retirement & AI Psychosis
1:00:34 Anthropic Study: Does AI Make You Dumber?