HukHukHuk, server #2, about 6:30am GMT+10, 18th of September

[Winterous]

the guy was still suffocating to death at the time.

Ah, I see now that I misinterpreted the first post as him having just died, that is quite embarrassing. I still think the laws should be adjusted to return the inaction clause though is that were the case, seeing as that is what this debate boils down to.

I think the inaction clause has too far-reaching effects on gameplay; it means the AI has to bolt security from getting to a traitor so they don't beat him to death, it means the AI has to bolt the doors to their upload so that laws which demand harm too (what are currently considered) humans don't get uploaded.
It demands an unreasonable amount of caution, and it's not good for gameplay.

I think that a reasonable wording would be something like...

Code:

You cannot, through action, cause a human to come to harm, or allow one to be harmed.

This would remove the ambiguity in my situation; my action was to stop dragging him to safety, which allowed him to come to harm; I maintain that it did not CAUSE harm since he was already subject to it (and in fact the position I left him was less harmful than where he was at first), but it definitely was ALLOWING harm.

Admins, what are your thoughts on that wording change?