Comment on An AI Just Attempted Murder... Allegedly... by SomeOrdinaryGamers [21:15 min] Video
thingsiplay@beehaw.org 1 day ago“Intent” is not that well defined. In example in Germany if someone drives drunken and as a result someone gets killed by it, then defendant (the person who drove the car) is accused of “intent to murder”, even if that was not the intention at all. Neglibility can cause intent.
So if the creators AND users of the LLM do not care about the results of people getting killed as a result, does it make them murder? Off course the LLM isn’t the murder here, I mean that’s without saying. It’s the human who is responsible.
MotoAsh@piefed.social 13 hours ago
That’s why it’s even more important to realize the machine has no intent. Its actions are solely the result of its creator’s actions in creating it.
I point out anthropomorphization so much because not only will it innoculate people against the advertising for it that WILL anthroporphize it, but when it fucks up, the appropriate people will be punished.
This isn’t a thinking machine going postal. It’s a dangerous product being pushed out with little regard for consequences.
Selling dangerous products used to mean something before billionaires bought the government…