[JimmyBuckets]

This seems like a great idea. Even without the linked surveys. Two questions I have:

- how you does this handle the fact that a lot of accounts on social media platforms are bots that maybe controlled by a small number of people.

- how do we actually get this implemented?

[63stack]

This is my question as well, especially about the "community check". How will it be ensured that the "community check" is not going to be dominated by bots pushing an agenda? How is that different from "just another comment section hidden behind a green button"?

[vintermann]

I guess they think fighting bots is a separate problem. Fair enough, since bots would still be a problem even if they pushed "reasonable" takes.

[energy123]

Regulating the algorithm is my favorite answer. Ban the recommendation engine on large social media sites. Make it a chronological feed of who you follow. Make it boring. I don't know all the details, but something like this.

[rcxdude]

I agree, recommendation algorithms are a huge part of the problem. Consciously choosing what you interact with is a very important part of media consumption IMO and most social media sites give you very little tools to do that (no, having likes/dislikes affect your personalised feed is not enough, especially when that also becomes 'engagement' and boosts it everyone's feed in general). These algorithms should be dumber in all areas except spam prevention (and even then, if there's less stuff in your feed you didn't specifically choose to see, spam should be a much smaller problem anyway).

[tskj]

I think this undersells the problem of discovery a little bit though. For example, youtube has been great at serving me longform content I want to engage with and wouldn't have discovered any other way.

(then they started having shorts, so I cancelled youtube premium)

[pibaker]

How many people do you think use HN's "new" and "comments" pages? These are exactly what you asked for, a feed of the website you are currently on without the recommendation system's influence.

I personally find them nigh unusable because of the lack of any kind of filtering. I am on HN precisely because it has a somewhat working post sorting system — a recommendation engine, as an activist who wants to get HN in trouble might say.

Anyway I doubt a regulation as such would fly under the first amendment. Recommendation is expressing and opinion and expressing an opinion is speech. If I think one post is better and deserves to be on the top spot, I believe I should have a right to say it without some guy in DC telling me to shut up.

[tardedmeme]

I want some kind of algorithm though. If some of my friends post a lot and some post a little, I want to see a more even split. And I want to see some posts from friends of friends, and from strangers who are posting similarly to my friends.

[plewd]

I honestly don't think it's possible for platforms to have "nice" algorithms like this without slowly slipping into the "maximum-engagement" algorithms we're plagued with now. I remember seeing this happen with Instagram, slowly going from a chronological feed to a confusing one where you can never be certain you've caught up with your network.

In a perfect world it would be great to have a platform that allows open-sourced algorithms for people to choose from, although that's a crazy pipe dream.

[johnpaulkiser]

I think tweaking section 230 in the US would have a similar effect. Make corporations liable for posts that they algorithmically amplify. The "discovery" algorithms would become banal overnight without an outright ban.

[bell-cot]

But as with pretty much every cardinal sin of late-stage capitalism - there are a whole lot of very entitled people, who are both very accustomed to and skilled at getting their own way, who are heavily invested in opposing any real solution to the problem.

[robot-wrangler]

> how do we actually get this implemented?

Hackers might be interested to know that there's an "open questions" section at the end of TFA. Some of it probably wants simulation, some wants theorems.

Camel-ai pubs/frameworks might be related and useful, for example: https://github.com/camel-ai/agent-trust

Several model checkers also have primitives for working with common-knowledge. TFA puts it like this:

> Learning a fact changes what you know. Seeing it displayed publicly — where everyone else can see it too — where you know others can also see it, changes what everyone knows, and subsequently how they act.

An important piece of technical vocabulary, it really seems we need this to talk about a lot of problems lately. Here is Terence Tao talking about some related math for disinformation and politics ( https://mathstodon.xyz/@tao/114866548969775485 ) and summing it up this way:

> we barely even have the vocabulary to discuss, let alone analyze, games in which control of information is a major battleground.

He kinda means in general though I think.. probably we can find heuristics and crunch a case or two