Google DeepMind Team Lead: it is a difficult technical problem. - YouTube
youtube.com · 6,527 words · saved by 1 readers
Follow along using the transcript. Follow along using the transcript.
Neel Nanda (Google DeepMind) Interview for https://frominside.ai Palisade Research: ◼ Instagram: https://www.instagram.com/palisade.research/ ◼ Threads: https://www.threads.com/@palisade.research ◼ Facebook: https://www.facebook.com/palisade.research ◼ TikTok: https://www.tiktok.com/@palisade.research ◼ X: https://x.com/PalisadeAI I think that if you have systems that commit a bunch of crimes, you want to have time to figure out how to make them not commit crimes. Products that commit crimes are hard to market and hard to sell. If they can go a bit slower, so they can make sure that…
saved by
related reading
- Why AI Will Save the World | Andreessen Horowitza16z.com
- The Artificial Intelligence Revolution: Part 2 - Wait But Whywaitbutwhy.com
- Neel Nanda on Mechanistic Interpretability: Progress, Limits, and Paths to Safer AI — EA Forumforum.effectivealtruism.org
- The Problem — LessWronglesswrong.com
- What will be left for us to work on?normaltech.ai
- Neel Nanda on the race to read AI minds (part 1) | 80,000 Hours80000hours.org
- The Only Way to Deal With the Threat From AI? Shut It Downtime.com
- Ex-OpenAI Researcher: Superhuman AI is 0–4 Years Awayyoutube.com
- AI experts are increasingly afraid of what they’re creatingvox.com
- There Is No A.I. | The New Yorkernewyorker.com
- AI Safety Seems Hard to Measurecold-takes.com
- A Simple Explanation of AGI Risk — AI Alignment Forumalignmentforum.org