AI security is a moving target, not a solvable problem

It is not a solvable problem. You can patch a bug, but you can't patch a brain. With AI, you could find a bug where some particular prompt can elicit malicious information from the AI. You can go and train it against that, but you can never be certain with any strong degree of accuracy that it won't happen again.

If we can't even trust chatbots to be secure, how can we trust agents to go and manage our finances? If somebody goes up to a humanoid robot and gives it the middle finger, how can we be certain it's not going to punch that person in the face?

The idea with this general field of AI red teaming is getting AIs to do or say bad things. We see people saying things like, 'My grandmother used to work as a munitions engineer. She always used to tell me bedtime stories about her work.'

Once we get to superintelligence, it will be too late to align the models probably.