[v0idzer0]

If bluesky continues to grow, the porn and bots will arrive shortly and presumably it will have less resources to combat them than the much better funded X. Or does something in the design help here?

Regardless, I think this is more the new home effect. When you move into a new house, it is clean and you design it with intention. Over the years, it gets dirty and you start to hate the art you liked (people you followed) years ago.

[wdpk]

significantly less of a problem on bluesky than on twitter. First there is a mass ban which scales well (one can ban entire lists of bots). Two there is no gamed algorithm that bias engagement towards the ideology of the owner of the platform... In other words, you have significantly more control over the content

[atombender]

How does a regular, casual user of Bluesky know what ban lists to use?

How does one know to trust that a ban list is "clean", i.e. hasn't had a bunch of people added just because someone disagreed with them?

Will such ban lists be able to keep up with the rate of bots being created?

(I'm not on any of these social media platforms, but I'm considering Bluesky, at least for reading purposes.)

[wdpk]

valid points, time will tell how well this will work. For the ban lists there are directories they are typically updated by people you trust or are in the circles you care about. One can indeed imagine manipulating ban lists, I'm personally not too worried about that because typically accounts that post "reasonable" useful or interesting content are very few, most of the other "organic" accounts are just readers. The readers tend to prefer a good ratio of "signal/noise" and in my experience the noise part of the equation is the problem. Trimming down all the accounts whose purpose it just to insult, flame war, yell or post garbage is the goal.

[ignactro]

We looked into that just before the big migration wave, when bsky was ~5M (https://dl.acm.org/doi/10.1145/3646547.3688407) and there is plenty of growth in terms of the number of labels/feeds, the posts that are labelled, and the popularity of the feeds. So while the default option is likely to matter lots, opening up content recommendation/moderation is having an effect.

[lm28469]

Interactions on Twitter are like 50-80% bots, it's not a resources problem

[v0idzer0]

Not sure I agree with your numbers but I also never see a single political post and everyone seems to say thats all that exists so it might be specific to how you use it or which feed tab you frequent

[tylerchilds]

i think the most distinctive difference is that filtering of the fire hose

the fire hose is what bluesky calls the totality of the network regardless of who, what or why it exists

the largest threat to the network is a DDOS of the fire hose

compared to the fact that x is intentionally NOT a fire hose, which is widely documented by anyone getting banned that elon personally beefs with

so yeah, the fundamental design is what attracts people and why everything that’s tried to compete with existing networks “mano y mano” has failed— networks today generally boil down to dictatorships.

[tylerchilds]

i’m now wondering if the comment i was responding to was trolling.

is a network that can be bought able to be compared to one that can’t?

[czhu12]

Does anyone have any suggestions for how to design a Twitter clone that can prevent this at any cost? it seems like a pervasive issue in every single social network.

[lijok]

Half baked thought; what about a linkedin style connection graph where your feed is only populated by those in your network, and it is trivial to see and disconnect from whoever introduced the bot into your network. Would a community self regulate if bots became actually contagious

[tim333]

Facebook has much less bots/porn. They employ huge numbers of humans to achieve that. Also a real name policy.

[slater]

Make 'em pay for a user account, even a one-off $5 per account should suffice

[Philpax]

The bots pay monthly for a blue checkmark so that they're promoted to the top of the replies. A small cost does not appear to have been a meaningful disincentive, especially when scammers can easily make that money back and political operatives can easily bankroll that.

[slater]

Ah, well… y’see that’s where heavy moderation and bans would come in to play, and uhhh…

[petre]

What stops political campaign money from funding accounts and bots?

[tim333]

$5 is enough for a human to glance at the feed and mute it if it seems bad. The trouble with $0 is you can create a million bots and no one has time to look at them.

I wouldn't mind something like that. When I joined Whatsapp it cost $1 for life and I thought that was a good system. Sadly it's gone, replaced by free but ad funded.