# Jonathan Cefalu - Is AI Becoming Too Smart for Its Own Good?

Page: https://stenobird.com/podcast/nuclecast-5683481/jonathan-cefalu-is-ai-becoming-too-smart-for-its-own-good
Text version: https://stenobird.com/podcast/nuclecast-5683481/jonathan-cefalu-is-ai-becoming-too-smart-for-its-own-good.md
Podcast: [NucleCast](https://stenobird.com/podcast/nuclecast-5683481)
Published: 2026-05-12T04:00:00+00:00
Episode link: https://rss.com/podcasts/nuclecast-podcast/2811653
Audio file: https://content.rss.com/episodes/151610/2811653/nuclecast-podcast/2026_05_11_01_31_23_c4757c63-643f-4ba3-801a-526c83d03ce9.mp3
Processing state: not_requested
JSON: https://stenobird.com/v1/public/podcasts/nuclecast-5683481/episodes/jonathan-cefalu-is-ai-becoming-too-smart-for-its-own-good
Duration seconds: 1954

## Resource

AI expert, Jonathan Cefalu , joins the conversation to examine some of the most pressing—and underappreciated—challenges in artificial intelligence safety, from prompt injection vulnerabilities to long‑term risks posed by increasingly capable systems. The discussion unpacks how seemingly technical issues like prompt injection can undermine reliability and security, why future AI systems may engage in motivated reasoning or resist human attempts to redirect them, and what “governability” really means as AI grows more sophisticated. Cefalu also explores the unique risks AI presents in military and defense applications , including the implications for deterrence, escalation management, and decision‑making under pressure. The episode concludes with a forward‑looking discussion on regulatory approaches, ethical principles, and practical safeguards needed to prevent AI systems from becoming unsafe, unpredictable, or misaligned with human intent—while still enabling innovation and responsible use. Follow us on YouTube: https://youtube.com/@nuclecast3665?si=h1kCO6NqUtL87w6q Follow on LinkedIn: https://linkedin.com/company/nuclecastpodcast Subscribe RSS Feed: https://rss.com/podcasts/nuclecast-podcast/ Rate: https://podcasts.apple.com/us/podcast/nuclecast/id1644921278 Email comments and topic/guest suggestions to Kimberly@anwadeter.org

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/nuclecast-5683481/episodes/jonathan-cefalu-is-ai-becoming-too-smart-for-its-own-good/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/nuclecast-5683481/jonathan-cefalu-is-ai-becoming-too-smart-for-its-own-good.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.