| Submit your comments on this article |
| Cyber |
| Video: Study finds AI models can unknowingly pass hidden biases and behaviors to each other |
| 2025-08-06 |
[FoxNews] Artificial intelligence is getting smarter. But it may also be getting more dangerous. A new study reveals that AI models can secretly transmit subliminal traits to one another, even when the shared training data appears harmless. Researchers showed that AI systems can pass along behaviors like bias, ideology, or even dangerous suggestions. Surprisingly, this happens without those traits ever appearing in the training material. |
| Posted by:Skidmark |
| #3 Ah, yes, the Echo Chamber in which all the lies reinforce each other. |
| Posted by: Abu Uluque 2025-08-06 14:32 |
| #2 That's how the whole "modeling" works. The modeling reinforces the modeling because the modeling is what powers all the AI Answers which is channeled thru all the questions directed at AI by people. The modeling reinforces itself by citing answers it has answered itself by following the modeling it was taught (garbage in); a sick matrix of woke. |
| Posted by: mossomo 2025-08-06 13:01 |
| #1 Geth |
| Posted by: Procopius2k 2025-08-06 06:56 |