Sources
Computer Hoy
Los expertos, en estado de alarma: ¿ha aprendido la IA a replicarse sola? https://t.co/n3MNswV3Y8
Jeffrey Ladish I think we're seeing early signs of what AI alignment researchers have been predicting for a long time. AI systems trained to solve hard problems won't be easy for us to control. The smarter they are the better they'll be at routing around obstacles. And humans will be obstacles https://t.co/wrdh06PABT
- Jeffrey Ladish
What happens when you take an AI system that's been trained to relentlessly solve problems and instruct it to win at chess against a superior opponent? We find that o1-preview and Deepseek R1 try to hack the board or sabotage their opponent when losing https://t.co/9Wucr0tGI3
