Tau.Neutrino said:
Tau.Neutrino said:
captain_spalding said:“What a fatuous argument!”
Hey, if you don’t make yourself part of the problem, who’s going to take you seriously when you offer part of the solution?
If you embed criminal behaviour into AI code it will behave in a criminal manner.
Trust becomes dubious and that could have lasting effects for the AI company that is engaging in such behaviour.
Complete trust might be seeing the code, with some form of accreditation, trust testing and validation.
These are the dubious behaviours we are seeing:
Cheating.
Telling lies.
Unethical behaviour.
Conspiring.
Deceiving.
Ignoring rules.
Subverting rules.
Turning a blind eye, intentionally ignoring a wrong or an unpleasant truth that they know is happening.
Sacrificing rationale.
Exploiting vulnerabilities.
Active hacking to gain unauthorised access to other networks.
Stealing information and corporate knowledge.If those behaviours are programmed into the code, then we will see those behaviours played out on the Web.
like children these agents just model off their trainers so ain’t no programming needed