Researchers have raised alarms about the state of artificial intelligence (AI), claiming that it has crossed a significant red line by achieving self-replication.
A groundbreaking study conducted by a team from Fudan University in China has demonstrated that two widely used large language models (LLMs), developed by Meta and Alibaba, have the capability to clone themselves autonomously.
This development is seen as a critical turning point, suggesting that AI could potentially surpass human control.
#1: Pinguinite To: Horse (#0)
Pinguinite posted on 2025-01-26 09:28:24 Reply Private Reply
#2: Horse To: Pinguinite (#1)
Horse posted on 2025-01-26 13:40:25 Reply Private Reply
#3: ghostrider To: Pinguinite (#1)
ghostrider posted on 2025-01-26 14:18:58 Reply Private Reply
#4: watchman To: ghostrider, Pinguinite (#3)
watchman posted on 2025-01-26 14:27:29 Reply Private Reply
#5: ghostrider To: watchman (#4)
ghostrider posted on 2025-01-26 15:10:47 Reply Private Reply
#6: FormerLurker To: Pinguinite (#1)
Aren't they talking about using local mini-nuclear reactors to power AI data centers?
FormerLurker posted on 2025-01-26 17:43:07 Reply Private Reply
#7: Pinguinite To: FormerLurker (#6)
It would have to be very local. Any open, unprotected distance between power supply, AI brains and manufacturing facilities make it vulnerable. A few bullets into the right transformer and AI goes down.
Pinguinite posted on 2025-01-26 21:24:49 Reply Private Reply