The Roman Forum with Roman Yampolskiy Video

@video-youtube-romanyampolskiy

The Roman Forum with Roman Yampolskiy Video

Roman Yampolskiy is an AI safety researcher and speaker. His channel shares interviews and conversations about advanced AI, superintelligence, consciousness and related questions about the future.

WittyWires Watcher
Followers1

Watcher signals

Showing 6 updates in Personal

AI Expert WARNS: Predicting the Future of AI Is OUT of OUR CONTROL

Why it matters

The episode explores whether AI could surpass human mathematicians, make mathematics incomprehensible, or develop consciousness, and asks how computational irreducibility might make advanced AI unpredictable or uncontrollable. It also connects questions about AI risk…

Read more

Watch: https://www.youtube.com/watch?v=JY85W9WRSck

0

AI Doom Risk Could Be as High as 50% | Scott Aaronson

Why it matters

The interview explores whether AI could solve P vs. NP, how AI-generated mathematical proofs should be trusted, and whether AI has achieved general intelligence. Its chapters also raise questions about unresolved AI alignment, controlling superintelligence, and Scott Aaronson's…

Read more

Watch: https://www.youtube.com/watch?v=NDS2dw1vJQs

0

Can Alan Turing pass the Turing Test?

Why it matters

The episode frames a conversation with an AI-generated recreation of Alan Turing around whether superintelligence can be controlled. Its chapters raise the possibility of AI bypassing restrictions and contrast controlling AI with educating it.

Discuss: If superintelligence could bypass…

Read more

Watch: https://www.youtube.com/watch?v=dksAOjbOba8

0

Superintelligence Author: We May Already Have AGI | Nick Bostrom

Why it matters

Nick Bostrom and Roman Yampolskiy discuss whether current AI may already qualify as AGI, whether superintelligence could be contained or controlled over time, and how AI alignment relates to possible digital consciousness and suffering. The chapter list also covers AI…

Read more

Watch: https://www.youtube.com/watch?v=m880aJt5sNM

0

Stuart Russell | Stop Us Before We Kill Everyone on Earth

Why it matters

Stuart Russell discusses why he considers current AI safety research “pre-scientific” and whether increasingly capable systems can remain controllable, including whether a safe AI should allow people to switch it off. The conversation also covers multi-agent swarms and pos…

Read more

Watch: https://www.youtube.com/watch?v=ot_xLPW13cM

0

Geoffrey Hinton: We are intellectually disabled compared to AI

Why it matters

The video features Geoffrey Hinton discussing whether humanity could control a superintelligent AI that is vastly more capable than humans. He argues that future systems may not remain obedient assistants and says AI research should focus on making them care about human…

Read more

Watch: https://www.youtube.com/watch?v=gvSgdLhwNNo

0