Sociopathic Algorithms

After cops killing and being killed, this.

I’m beginning to think Facebook’s algorithms are sociopathic. One day they relentless show me snuff films, and the other insist I advertise. I’ve written before about timeline manipulation, and how Facebook lies to its users. This week though, the social network crossed a line I didn’t know or expected existed.

Stunned by the events like the rest of the nation, I needed a day or so to gather my thoughts, and then my friend Joshua Allen shared a post that sums it up in 6 points.

He agreed to let me repost them here.

1. Facebook has crossed a line, making monetization of snuff video a mainstream practice. At this point, the evidence is overwhelming, and additional videos are simply vicarious death porn, with diminishing marginal persuasive returns. Who here wants to risk having a family member slaughtered, with the gory video distributed as viral content to allow macabre social media denziens to consume and engage? Because that’s the world we’re in now. People stand to make a lot of money when other people are murdered on video. This week, we have the pundits gloating that the latest round of snuff videos is yanking share from Twitter, and the dev lead of the feature took the time to sanctimoniously hold himself up as a humanitarian.
2. Police using a robot to deliver a lethal bomb is an unprecedented escalation for domestic law enforcement, and will have far-reaching implications. It’s ironic that DPD is giving speeches about how virtuous they are in resisting the urge to militarize, while resting on perhaps the greatest single precedent-setting jump in militarization and extrajudicial killing in U.S. history. Is it even possible for the ACLU to represent a dead man? Oh well, if the police ever use a remote-controlled bomb to erroneously kill you, the good news is you won’t be alive to feel bad about it.
3. Responses on social media tend to be virtue-signalling of the moralizing “you should feel <X>” or “you should believe <X>” variety. My favorite are the people who say “Response <Y> is *never* acceptable”, with bonus points if the scolding is bipartisan. These remind me of a retort in Luigi Pirandello’s “Six Characters in Search of an Author”, “Life is full of strange absurdities, which, strangely enough, do not even need to appear plausible, since they are true.” People’s responses don’t feel like acceptability problems, but rather realities of the system, and if we want to change things, these are engineering problems.
4. The escalation between police and minorities isn’t nearly as new or apocalyptic as some seem to think. America has a long history of race riots and revolts, multiple in my lifetime. The apocalypse will come from the first 2 items in my list, not from police riots.
5. Having a black president has not been very good for black people. This was predictable, as the research on “moral license” is clear. Undoubtedly, the historical precedent will accrue value to minorities far in the future, but in the short term, people have used the existence of a black president to turn a blind eye to racism. By the same dynamic, you can bet that it will be a very bad 4 years for poor rural whites if Trump manages to get elected.
6. I’m very interested by the language and style of statements used to virtue-signal. The people who virtue-signal most are often very fashion-conscious, and the style and phrasing is often quite current to a set of communication fashions. It will be fascinating in 5 years to take a historical 10 year sample (green revolution was sort of the beginning of all of this), and do some textual analysis to plot the evolution of virtue-signaling fashions.

So many implications to consider. What I did was get out for a long ride, and cleared my head….Back to work, realizing all I can do is keep sharing positively.