EPISODE · Nov 14, 2025 · 18 MIN
Is GPT-5.1 Really an Upgrade? But Models Can Auto-Hack Govts, so … there’s that
from AI Explained Official Podcast · host Philip - Host of AI Explained YT
A lot just got released in the last 36 hours, and it will all affect hundreds of millions of people. 10 details you would miss if you just read the headlines, from GPT 5.1 regressions, to how Claude hacked Govt Agencies, to SIMA 2, and Musical Turing Tests.https://assemblyai.com/aiexplainedChapters:00:00 - Introduction00:56 - GPT 5.1 Smarter?01:47 - Some Regressions03:22 - Sycophancy?05:22 - Claude Auto-Hacking 06:16 - Jailbreaking through Granularity08:22 - This Will be Re-used09:30 - Hallucinating Hacker09:57 - Surprisingly Neutral Tone12:18 - SIMA 214:10 - Alpha Parallels17:24 - AI MusicGPT 5.1 Announcement: https://openai.com/index/gpt-5-1/System Card: https://cdn.openai.com/pdf/4173ec8d-1229-47db-96de-06d87147e07e/5_1_system_card.pdfBenchmarks: https://openai.com/index/gpt-5-1-for-developers/Simple Bench: https://lmcouncil.ai/benchmarksAuto-Hacking: https://x.com/AnthropicAI/status/1989033793190277618https://www.anthropic.com/news/disrupting-AI-espionageReport: https://assets.anthropic.com/m/ec212e6566a0d47/original/Disrupting-the-first-reported-AI-orchestrated-cyber-espionage-campaign.pdfSima 2 Announcement: https://deepmind.google/blog/sima-2-an-agent-that-plays-reasons-and-learns-with-you-in-virtual-3d-worlds/https://x.com/amoufarek/status/1988986075331858693Scepticism: https://www.technologyreview.com/2025/11/13/1127921/google-deepmind-is-using-gemini-to-train-agents-inside-goat-simulator-3/Voyager: https://voyager.minedojo.org/Reuters Music: https://www.reuters.com/legal/litigation/are-you-listening-bots-survey-shows-ai-music-is-virtually-undetectable-2025-11-12/
Embed this episode
What this episode covers
A lot just got released in the last 36 hours, and it will all affect hundreds of millions of people. 10 details you would miss if you just read the headlines, from GPT 5.1 regressions, to how Claude hacked Govt Agencies, to SIMA 2, and Musical Turing Tests. https://assemblyai.com/aiexplained Chapters: 00:00 - Introduction 00:56 - GPT 5.1 Smarter? 01:47 - Some Regressions 03:22 - Sycophancy? 05:22 - Claude Auto-Hacking 06:16 - Jailbreaking through Granularity 08:22 - This Will be Re-used...
Ready to play
Is GPT-5.1 Really an Upgrade? But Models Can Auto-Hack Govts, so … there’s that
No transcript for this episode yet
Similar Episodes
No similar episodes found.