On a different topic but answering to the same quote : advancements in quantization of models to significantly reduce model memory consumption for inference without reducing model performance might also mitigate the imbalance between ALU ops and memory bandwith. This might only shift the problem a few orders of magnitude away, but still, I think it‘s worth mentioning.
It wasn’t possible to reach full immunity but you never need to be the fastest gazelle to escape the cheetah, just faster than the slowest one.
and
intimate relationships and acquaintances are the number 1 group of perpetrators
The "faster than the slowest one" doesn't apply in the context of sexual violence. You might not care about someone else's bike being stolen but you will never drink less hoping your friends or others are assaulted instead of you.
In the defense of oral conversations (especially in person), any conversation dealing with emotions is much easier for me in person. There is so much to be wrongly interpreted through text : an emoji with many possible interpretations, the meaning of a dot at the end of a sentence, guessing the actual emotion of the other person behind the text… I think there are many ways to misinterpret a text and the amount of text required to make sure your message is correctly understood is often greater than if the person was in front of you + it can be annoying (for... (read more)