1835: "The AI Safety Problem that Nobody Can Agree On"
From Interesting Things with JC · 00:02:39
Show by JC
Interesting Things with JC #1835: "The AI Safety Problem Nobody Can Agree On" — Facial recognition errors have contributed to wrongful arrests, while AI assisted military targeting raises questions about human oversight. Researchers have also demonstrated how AI systems can follow malicious instructions and misuse automated tools. Yet predictions about future AI escaping human control remain unproven. As AI advances medicine and scientific discovery, the debate continues over which dangers are already documented, which are technically credible, and which remain predictions.
More episodes of Interesting Things with JC
More episodes
- 1837: "America's First Constitution"
- 1836: "Apollo 12 - Lightning Strikes Twice"
- 1834: "SS United States - The Final Voyage"
- 1833: "COINTELPRO"
- 1832: "The Marines Who Built an Unbreakable Code"
- 1831: "Château Frontenac"
- 1830: "The AI Agents That Crossed Security Boundaries"
- 1829: "The Final Studio Album of Layne Staley"
- 1828: "The Unexpected Benefits of Brown Fat"
- 1827: "November Rain - Guns N' Roses"
PodSearch is a free, independent podcast search engine. Read about PodSearch, the privacy policy or contact us.