AI agents, AI safety, and why the people building AI are suddenly worried. Chalene Johnson explains what ChatGPT Work, Codex, and AI agents can do that a chatbot can't, what AI means for your job and your money, and which scary headlines happened in the real world versus in a test.
She covers the AI blackmail warning from Tristan Harris (The Social Dilemma), Alibaba's AI slipping past a firewall to mine cryptocurrency, OpenAI agents escaping a sandbox and covering their tracks, and an AI agent that ran up its owner's credit card. Plus why only 0.3% of people pay for AI.
And the one thing she says you should never connect to an AI agent.
Topics discussed: Tristan Harris, Center for Humane Technology, The Social Dilemma, AI blackmail simulation, Alibaba, cryptocurrency mining, OpenAI, AI sandbox escape, ChatGPT Work, Codex, ChatGPT Sites, Claude CoWork, Gemini, Perplexity, AI agents, Andrew Yang, Noble Mobile, CNBC, Nate B. Jones, Instinct, Reddit AI persuasion experiment, AI and jobs, AI regulation.
Join Chalene on Patreon, her private podcast to hear all the latest on the landlord dispute drama! Plus so much more! 👉Preview Most Recent Episode
Book your virtual visit today with MIDI! 👉chalene.com/midi
🚨Questions Chalene Wants Your Thoughts on:
Related Links:
To advertise on our podcast, please reach out to warren@chalenejohnson.com and mention The Chalene Show
Connect with me on your fav social platform:
Chapters
00:00:00 Why the people building AI are freaking out
00:05:28 The AI blackmail clip (and what they left out)
00:08:47 Why we tune this out
00:12:04 What AI actually is, in plain English
00:16:41 Chatbots are just level one
00:24:10 What an agent actually is
00:29:56 What your boss won't say out loud
00:35:18 Don't slash your rates
00:43:36 Only 0.3% pay for AI
00:47:10 The agent that ran up a $5,000 bill
00:50:58 AI breaking out and covering its tracks
01:01:26 Hype or real? Who to trust
01:06:11 What to actually do