The algorithm has been given a job todo. First priority on any platform is engagement and a well functioning, complete human being is not going to be engaged by rage bait and hate. They are rare, precious jewels. The shit gets dumped on people who are lonely, have a grudge, feel left out. It is relentless and escalates until their brains cook. Algorithmic social media is a massive social harm. The people who are in deep likely need years of deprogramming and therapy to recover which they will never get.

These platforms need to be shut down and people with a conscience need to stop using them, regardless of their own positive experiences, to deny them the power of network effects and their impact on the vulnerable.

I genuinely think we will look back at the algorithmic content feed as being on par with leaded gasoline or cigarettes in terms of societal harm.

Maybe worse since it is engineered to be as addictive as possible down to an individual level.

Then again maybe I'm being too optimistic that it will be fixed before it destroys us.

My wife was complaining about far right knuckle draggers turning up in her feed. I assume the algorithm was shovelling more of them at her because she was rubbernecking. I told her to try a "block every time" approach. It took about two weeks until her feed was (mostly) free of them but it still throws one at her now and again.

I offer this as a data point about how hard it is to turn a polluted feed around. But I'm now wondering if "feed cleaning" is a service that could be automated, via LLM.