We're sunsetting PodQuest on 2025-07-28. Thank you for your support!
Export Podcast Subscriptions
back
“What’s the short timeline plan?” by Marius Hobbhahn
44:22
Share
2025/1/2
LessWrong (30+ Karma)
AI Chapters
Transcribe
Chapters
Why Are Short Timelines Plausible?
What Are the Minimum Achievements Needed?
Conservative Assumptions for Safety Progress
What’s the Plan?
Layer 1: Keeping a Faithful and Human-Legible CoT
Significantly Better Monitoring
Control Without Assuming Human-Legible CoT
Deeper Understanding of Scheming
Evaluations and Security
Layer 2: Improved Near-Term Alignment Strategies
Continued Work on Interpretability and Oversight
Reasoning Transparency and Safety Culture
Known Limitations and Open Questions
Shownotes
Transcript
No transcript made for this episode yet, you may request it for free.
