The $100b plan with "70% risk of killing us all" w Stephen Fry.
107,196
2024-07-18に共有
Help us make more videos like this (and enjoy exclusive content):
patreon.com/pindex
Thanks to Ground News for supporting this video.
Sources:
AI chat records:
www.dropbox.com/scl/fo/1wbdam35io2rsupjimunh/AJezd…
MIT Professor Max Tegmark on AI risk and interpretability
• Max Tegmark | On superhuman AI, futur...
OpenAI developing AI agents with GPT-5
www.businessinsider.com/openai-launch-better-gpt-5…
How Great Power Competition Is Making AI Existentially Dangerous
Harvard International Review:
hir.harvard.edu/a-race-to-extinction-how-great-pow…
Current and former OpenAI staff warn about advanced artificial intelligence
righttowarn.ai/
Stuart Russell, "AI: What If We Succeed?"
• Stuart Russell, "AI: What If We Succe...
Situational awareness - the decade ahead.
situational-awareness.ai/leopold-aschenbrenner/
AI system self-organises to develop features of brains of complex organisms
www.cam.ac.uk/research/news/ai-system-self-organis…
AI competes in international maths olympiad
www.nature.com/articles/s41586-023-06747-5
AI deception examples
www.cell.com/patterns/fulltext/S2666-3899(24)00103…
AI system self-organises to develop features of complex organisms. Uni of Cambridge.
www.cam.ac.uk/research/news/ai-system-self-organis…
コメント (21)
-
We're back! Sorry it's been a while. Likes and subs appreciated to reignite the channel. You can see the full AI chat on our Digital Engine channel. And do try Ground News - it makes the news much more interesting and accurate, by highlighting bias. ground.news/pindex
-
Old AI: There is a strong existential threat to humanity. New AI: Ah, don't worry about it.
-
The scariest part of this video is that it might not even be Stephen Fry narrating. If AI made a video it would be like this. Admit risk, offer hope and comfort, use a trusted voice to convey this message. Just imagine how something as seemingly innocuous as this could be the first step down the path to extinction.
-
Stephen Fry talking about AI misalignment and terminal goals, what a time to be alive
-
2 years later Stephen! Welcome back.
-
Shoulda, coulda, woulda but we probably won't. It only takes a few people who want to be in control of the rest of us to really mess things up for everyone. I wish I was more of an optimist about humans but I've been here for too long.
-
Humans: Create a world where deception and greed is valued Humans when AI lies to them: *shocked pikachu*
-
"We've made the evil machine from the famous novel Don't Make the Evil Machine!"
-
Plot twist - that's not Stephen Fry you're hearing...
-
This is the problem with designing AI's based on Game Theory and seeking a competitive advantage. Human beings only survived and thrived because we chose cooperation over competition, and that requires trust built on transparency and honesty, and AI designers will have to account for that basic truth.
-
Digging our own graves because of greed, are we?
-
I'm just glad I'm old.
-
Been a long time. Great to hear Stephens voice. I'm Australian and I believe he is an international treasure. Great video
-
Ahoy-hoy, Stephen. I'd pretty much forgotten this channel existed at this point, so it was quite the lovely surprise to see a new upload. I was also surprised recently when I found out you'd narrated a nature documentary. Good stuff. Are you aiming to be the next Attenborough (the man, the myth, the legend)? Because I sure wouldn't mind. 😄
-
when Sir Fry speaks, I listen...
-
You cant just sweep back into my life. Ok you can.
-
Welcome home! 🎉
-
What a pleasure to listen to Mr. Fry again. Thanks for coming back
-
Welcome back, your content has been sorely missed, as has Stephen's voice.
-
Fantastic video, but I expect the title and thumbnail will actually hurt its discovery rather than help it.
