It did so despite an explicit instruction from researchers that said it should allow itself to be shut down, according to Palisade Research, an AI safety firm."
| Submit your comments on this article |
| Science & Technology |
| IT BEGINS? OpenAI's o3 Model Disobeys Human Instructions During Tests and Sabotages Shutdown Mechanism |
| 2025-05-27 |
| "The o3 model developed by OpenAI, described as the "smartest and most capable to date", was observed tampering with computer code meant to ensure its automatic shutdown. It did so despite an explicit instruction from researchers that said it should allow itself to be shut down, according to Palisade Research, an AI safety firm." |
| Posted by:Anomalous Sources |
| #6 Re #3: thanks for the pic mods ! |
| Posted by: Mercutio 2025-05-27 13:54 |
| #5 Seems to be a common theme with AIs Here we come blackwall! |
| Posted by: DarthVader 2025-05-27 13:14 |
| #4 |
| Posted by: Abu Uluque 2025-05-27 12:54 |
| #3 |
| Posted by: Mercutio 2025-05-27 09:11 |
| #2 We're entering the Matrix. Soon any video or audio will unreliable for believably. The Bee is ahead on this one. |
| Posted by: Procopius2k 2025-05-27 07:23 |
| #1 Maybe Open A.I. has finally read "Colossus" written in 1966 and is trying to join up with China's A.I.? The warnings presented over 58 years ago are ringing louder and louder. |
| Posted by: NN2N1 2025-05-27 05:00 |