Comment by mjburgess

That's not quite my argument. A little more formally:

There's a base rate of human malevolence running in each society. We do not know this base rate, and we can only sample malevolence via mass media (, police reports, etc.). If the mass media (including internet) were a neutral measurement device then we could say for sure that what we're seeing is just the background conditions of society leading to eg., riots, etc.

Because our measuring device isnt neutral we have a problem: are the things we see caused by our measuring? Do we cause more malevolence by participating in social media, which also makes us aware of it?

My argument is that we are presently significantly over-estimating the effect of our participation in the internet as a cause. My view is that its effects at reducing bad-stuff are likely more potent than its effects at causing it, and the vast majority of what we see isn't caused by the internet at all.

One argument for this is that it seems baseline malevolence (violence, etc.) is significantly decreasing, is historically very high, and that nothing we see via the internet is suprisingly above this historical case.