[Question] Best giving multiplier for X-risk/​AI safety?

[SiebeRozendal]

What do you currently consider the best giving multiplier for X-risk/​AI safety? (highest & somewhat robustly supported by evidence)