4 ms·
To me, the problem currently I believe is with the internet and how it creates a new quality to speech. People can speak up, but it turns out that there's now
by didibus 4y ago
To me, the problem currently I believe is with the internet and how it creates a new quality to speech.
People can speak up, but it turns out that there's now too much noise and nobody can distinguish the source of the speech, the context of the speech, the trustworthiness of the speech, and the occurrence of the speech (how many people are saying the same thing versus parroting what they were told).
Not only that, but the cost to speak is so low now, that bad actors have a field day, and I include bad faith people looking to just amass political fame, power, following, etc.
Mix that in with our new shortened attention span, and people no longer engage with the speech, instead it becomes like advertisement, if you've seen the same message a lot you'll subconsciously have taken it as a truth. It becomes which opinion have you seen most becomes your opinion, and everyone does this, so the truth becomes a matter of who can parrot their message as often, as loudly and to as many people.
This also means that catchy speech will win over non-catchy speech, because everything is more like a marketing game. Appealing to emotions is attention grabbing, controversy is attention grabbing, so all speech of that form becomes the winning speech in terms of getting viewed and heard.
That's why I think all reasonable people are doomed here, this mechanism of social conversation lends itself automatically to destabilizing opinions winning over stable reasonable but boring and not attention grabbing perspectives.
Before the internet, there were more filtering happening in a layered approach, publications and journalist were there to filter out the noise, the low quality, the unsubstantiated, the agenda driven, and all those.
Speaking up about injustices involved having to physically mobilize yourself to a public square, which takes a lot of effort from the speaker, again allowing a filtering out of bad actors and bad faith individual since the effort demanded had a cost, only people really motivated by great injustice would have the will to go through with protesting.
I don't think it was ideal before either, as publications as a filter had their own bias to what they'd filter out. But at least even what they let through, since there wasn't an abundance of it, the listeners could have a deeper think about it, and didn't believe it to be true just because they keep hearing about it over and over.
Hopefully we can find a way to improve on both. I think we need to put more contextualizing and filters over the internet speech, nothing that removes or bans speech, but allows to manage all the noise.
Algorithms are already in place doing this, except they filter based on corporate profit making, which often involves around engagement and attention grabbing. Again this is extra counterproductive, worse than a simple unfiltered approach with everything chronological.
I'm not sure how best to implement improvements, but I feel strongly it is needed, because even I can no longer make sense of all the noise on the internet, and it feels more and more like I actually can't tell what's real or not anymore, even though I'm in an ocean of information.