Meta said it based the $1.4 trillion figure, which is nearly as large as the company’s entire market cap, on how the attorneys general of California, Colorado, Kentucky and New Jersey have argued p…
And how do you quantitatively differentiate between “sorting posts” and “sorting posts, but evil”? Moving a post up the list is by definition prioritizing it. Moving another down the list is suppressing that one. Which posts get which treatment is governed at least in part by dopamine, since people are going to make pull requests based on how they want to use the site.
There are a couple of fundamental properties of the Fediverse that give it decently robust immunity against the “evil” shit that either has been, or will be, the downfall of all the traditional closed-source social media entities.
Because it’s FOSS, there’s transparency into the way internal things are supposed to work, and everything that happens on here, from mod actions down to the level of individual votes can be tracked and audited.
AFAIK, the actual sorting/ranking of your feed happens on the client side (on the device you’re browsing from). And is strictly for user quality of life. Because those are the views the bulk of users like to see. It may even differ depending on your client, but they seem to operate more or less the same.
The server-side has no reason to give different weights to individual posts/comms/users, nor any mechanism with which to do so. The incentive for someone to unilaterally decide what users see on this platform simply doesn’t exist… I’m sure it’s targeted by individual bad actors all the time, but the very nature of this platform makes it next to impossible for any one person or entity to gain enough leverage to rise above nuisance-level. There’s really nowhere on here to hide malicious intent long enough to accomplish a meaningful level of funny business before the network cuts you out or grows around.
The instance database of course stores the communities, the actual content of each post and comment, associates everything with the relevant users, keeps track of timestamps and the individual vote counts. Most everything beyond that gets layered on by whatever you’re using as a client.
Given the utter simplicity of the ActivityPub protocol and the ability of any interested party to gain granular visibility on everything down to the level of who voted which way on what comment? It seems to follow that anything falling outside the pattern of those parameters would raise immediate suspicion and get dealt with pretty quickly.
Advertisers and corporate types are just not particularly attracted here. The prevailing incentives just aren’t compatible with their aims.
I suppose you could say “well, what if the devs have a change of heart and go in a different direction, go rogue or start walling things off?” Whatever project that happened to would get forked the instant a whiff of impropriety was detected. The fork would probably end up better than whatever it replaced!
If a rogue instance started being a little too weird it would get dealt with and defederated just as swiftly by a large mass of federated instances, isolating it.
And how do you quantitatively differentiate between “sorting posts” and “sorting posts, but evil”? Moving a post up the list is by definition prioritizing it. Moving another down the list is suppressing that one. Which posts get which treatment is governed at least in part by dopamine, since people are going to make pull requests based on how they want to use the site.
There are a couple of fundamental properties of the Fediverse that give it decently robust immunity against the “evil” shit that either has been, or will be, the downfall of all the traditional closed-source social media entities.
Because it’s FOSS, there’s transparency into the way internal things are supposed to work, and everything that happens on here, from mod actions down to the level of individual votes can be tracked and audited.
AFAIK, the actual sorting/ranking of your feed happens on the client side (on the device you’re browsing from). And is strictly for user quality of life. Because those are the views the bulk of users like to see. It may even differ depending on your client, but they seem to operate more or less the same.
The server-side has no reason to give different weights to individual posts/comms/users, nor any mechanism with which to do so. The incentive for someone to unilaterally decide what users see on this platform simply doesn’t exist… I’m sure it’s targeted by individual bad actors all the time, but the very nature of this platform makes it next to impossible for any one person or entity to gain enough leverage to rise above nuisance-level. There’s really nowhere on here to hide malicious intent long enough to accomplish a meaningful level of funny business before the network cuts you out or grows around.
The instance database of course stores the communities, the actual content of each post and comment, associates everything with the relevant users, keeps track of timestamps and the individual vote counts. Most everything beyond that gets layered on by whatever you’re using as a client.
Given the utter simplicity of the ActivityPub protocol and the ability of any interested party to gain granular visibility on everything down to the level of who voted which way on what comment? It seems to follow that anything falling outside the pattern of those parameters would raise immediate suspicion and get dealt with pretty quickly.
Advertisers and corporate types are just not particularly attracted here. The prevailing incentives just aren’t compatible with their aims.
I suppose you could say “well, what if the devs have a change of heart and go in a different direction, go rogue or start walling things off?” Whatever project that happened to would get forked the instant a whiff of impropriety was detected. The fork would probably end up better than whatever it replaced!
If a rogue instance started being a little too weird it would get dealt with and defederated just as swiftly by a large mass of federated instances, isolating it.