Post #4383783

@mttaggart@infosec.exchange This is a very interesting perspective. My inital take on breaking the argument down.: Assume using AI does harm. Should be easy enough there. Water and energy usage, lowered earning power for people in many sectors, etc. Assume AI posiitively impacts some security function. Again, pretty easy. Even if it can be exploitrd with hidden instructions, etc. and isn't always reliable it is probably faster and more effective at immediate triage with a manual follow up at least. I could believe a benefit exists there. So then the moral argument is whether acceptable to contribute to an indirect mediated harm to many in order to provide a benefit that is more immediate to a smaller group of people.