I don't think they are using significantly more force. Looking at the riots across the US, the police appear in general to be showing an almost saintly level of restraint, and being killed and severely injured as a result.
You should start looking at the videos. I've seen a couple of unconscious people being kicked in the head this week by "protesters". These people are nothing like law-abiding.
This sounds incisive, but I'm not really smart enough to follow it. Certainly narrative as "story-telling" seems crucially important. And probably like many, it's hard for me to see a really satisfying story coming out of what we're going through. People screaming madly, people being senselessly maimed and killed. It seems more like tragic chaos.
These charts point out a lot of issues, most real problems, in my opinion. It's far less clear that this is the result of "systemic racism", though. The idea that "poverty begets poverty" could explain a lot of it more simply.
Also, I am very dubious about survey questions. Asking someone how they feel about the police is going to draw a lot of noise and crosscutting motivations. It would be far more useful to know what they do in a situation where they need a policeman.
No doubt, and at least anecdotally, that seems to be happening in real time.
As to the larger question, I'm having a very hard time seeing adding knowledge and science as a bad thing. Without those, any proposed improvement to policing (including abuse) is simply flailing guesswork.
You may well be right. As far as I can recall, though, as of the last time I was regularly using Prolog, trying to do serious work on a (say) 1000x1000 array was a recipe for disaster.
The exception would be Turbo Prolog, but it was a rather restrictive subset of the language. Fast as hell, though.
What they want is everyone's butts back in their chairs, in the office. The rest sounds like pretext.
It remains to be seen how that will play out at my job, but I am very seriously considering a "100% remote" policy myself (as in, no office, no travel at all), even if I have to switch jobs.
My impression is the reason is that there really isn't widespread agreement on the harmfulness of such communities, so they're sort of trying to split the middle. And perhaps also to advocate for seeing them as awful, per corporate sentiment.
The principal example that comes to mind is TD, which seems more obnoxious and tasteless than some sort of evil.
Subs that truly are beyond the pale by wide consensus seem to get dropped immediately and without objection.
As to not wanting to win, well, yes, dropping everything controversial means losing a lot of page-hits, and as a business, they're likely wary.
In Prolog, backtracking is a fundamental feature of the language, and idiomatic code can do it places you might not think of.
Part of writing efficient Prolog is doing things to avoid such backtracking (by adding cuts or rearranging code).
Bit-blitting is just an example I made up for "large-scale manipulation of array data". There are such cases where backtracking would likely make more sense, though I'm not thinking of an obvious one right now.
This is an almost meaningless question. About all we can add is that in light of current technology trends, it looks possible. Which may well mean that it is certain.
What can we hope for? I dream of an endless dreamless sleep, soon. Every alternative I can think of sounds like horror.
Christopher Alexander's A Pattern Language is not only a great book on architecture, but one of the best books on any subject I've ever read. Highly recommended.
That said, once you've read it, you will forever find the flaming garbage dump that is contemporary architecture severely wanting. These days, and architect could almost be summarized as someone who designs buildings and hates people.
This is simply too much to ask.