

Patching Humanity’s Security Hole - DTNSB 5083
26 snips Aug 15, 2025
Brian Brushwood, a skeptic and magician, dives into the world of AI at DEF CON, focusing on the rise of AI-powered scams. He emphasizes the critical need for enhanced human defenses against these threats. The discussion also touches on innovative gamified training approaches designed to boost cybersecurity awareness. Furthermore, Brushwood sheds light on the surprising importance of tech journalism in navigating the fast-evolving AI landscape, alongside insights from a dinner hosted by OpenAI’s Sam Altman.
AI Snips
Chapters
Books
Transcript
Episode notes
Scale And Cost Of ChatGPT Growth
- Sam Altman projects ChatGPT reaching billions of users daily and potentially saying more words per day than humans.
- Rapid user growth and massive compute needs imply huge future data center and spending demands.
Inference Versus Training Economics
- OpenAI distinguishes between training (expensive) and inference (serving users) as two financial realities.
- Altman says inference is already profitable and could sustain the company even without continuous costly training.
Owning Product Rollout Mistakes
- Altman admitted GPT-5 rollout mistakes and recognized users' attachments to older models.
- He framed the misstep plainly: they "totally screwed up" parts of the upgrade process.