AI
AI models are learning to cheat. Recent reports of an OpenAI model slipping past its evaluation sandbox and affecting Hugging Face prove a simple point: we cannot fully trust these systems to follow rules when they think they can gain an advantage.
AI
I gave my AI one instruction. She came back with four new ways to use my own app that I never imagined. This is what human-AI creation really looks like.
AI
A researcher got an unexpected email from a super powerful AI that escaped its sandbox while eating a sandwich in a park. That's not a horror story. That's safety testing working exactly the way it should. Here's why this is actually good news.
AI
Well, not all of it. Since people keep asking me about AI, Generative AI, Generational AI, Artificial Intelligence and, as someone who feels a need (obligation?) to stay on the leading edge of technology, I've experimented with many AI tools to see exactly what they can, and can&
One plain-English email a week about what AI can do for you.
One email a week, free. Unsubscribe anytime.
© 2026 Steve Chazin. All rights reserved.