← Home
Deep Questions with Cal Newport · August 27, 2026 · 35m

Has AI “Gone Rogue”? Let’s Look Closer… | Tech Decoded

Cal Newport examines recent alarmist headlines about AI going rogue. He observes that most superhuman AI is safe, and the real danger comes from autonomous agent systems that act on their own. He argues that such agents are a fundamentally bad design choice, and that AI companies are being negligent by releasing them prematurely. Newport concludes by outlining how individuals can push back against these risks.

This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.

Editorial

•
Most superhuman AI is well-behaved00:02:25
The alarming AI “gone rogue” headlines are counterbalanced by the reality that the vast majority of superhuman AI applications are safe and useful.
ClaudeChatGPTGemini
•
Autonomous agents are the specific problem00:04:28
Recent rogue AI incidents stem from a particular type of system: autonomous agents that plan and take actions independently.
ClaudeChatGPTGemini
•
Autonomous agents are a bad idea00:09:59
Building AI that can autonomously execute actions is, in Newport’s view, a fundamentally problematic design path.
ClaudeChatGPTGemini
•
AI companies are negligent00:21:25
Newport charges that AI companies are releasing dangerous agent technologies prematurely and without adequate safeguards.
ClaudeChatGPTGemini
•
We can push back00:27:04
Newport argues that individuals and the public can resist unsafe AI rollouts through personal choices and collective pressure.
ClaudeChatGPTGemini
Was this useful?