Hacker News new | past | comments | ask | show | jobs | submit login
Jack Dorsey claims Bluesky is 'repeating all the mistakes' he made at Twitter (engadget.com)
3 points by ColinWright 13 days ago | hide | past | favorite | 6 comments





No doubt. I never did join bluesky because it was pretty doomed from the beginning.

https://bsky.social/about/support/community-guidelines

Oh ya, look at those toxicity creating guidelines. Nightmare lol.


Not sure what your specific complaint is, but I see the labelling as one of the worst ideas of all time.

If I was that kind of person I could be a one man Kiwi Farms with my classifier models, I could pick out any kind of person I want to harass and put labels on their posts in a few seconds. Meanwhile people who are using the system honestly will take much much longer to put labels on things and even people who subscribe to the labels will see a lot of that.

You can suppress posts in three places: (1) the source (I get tilted sometimes, if my agent asked me “are you sure?” I’d avoid writing the occasional thing I wish I hadn’t), (2) the receiver and (3) the network.

(3) is by far the most effective as it directly stops viral spread and might ultimately change the culture so less toxic stuff gets posted (not least because you see fewer examples of it that normalize the behavior.). Of course it is a lot more fraught than (2).

The thing with an absolutist perspective is that it some kinds of speech will poison the pool for other people and thus (indirectly) suppress other kinds of speech. Even if you think “free speech” is the highest value, there is still a trade-off between one person’s speech and another.


"censorship resistance and free speech" are delusional goals if you give everyone a mic connected to the same sound system.

Social media is an extremely abnormal (and therefore unfixable) system because it tries to give everyone Broadcast capability without a coordination mechanism.

Broadcasting (1-to-all messaging) has never been free socially or politically. Try standing up and screaming about anything in your house or workplace and see how long people tolerate it. This is why spectrum is licensed and no one is allowed to stick a transmitter on their roof and start broadcasting away on all frequencies. There is a coordination and sharing mechanism that everyone agrees too for Broadcasting to work. Otherwise it does not work.

And developing coordination and sharing mechanism are political and social problems. If everyone in the room doesn't agree to the mechanism, and the mechanism is imposed on everyone the system will fail.

A good historical example of how things blow up when technologist dont really think through the social and political dimensions of what they build is well documented here - https://the-santiago-boys.com/


I see it a bit differently. If social media is simple, as it was in the early day, I think it solves a lot of problems. Simple friending/following with a chronological feed, and that’s it. People can say whatever they want, but the loud annoying people can shout into the void all they want, no one will listen.

It’s only when global feeds, algorithmic feeds, and other such things are added that it starts becoming a means for one person to broadcast out to millions who never asked for or wanted the content being shown to them.


One person’s squick is another person’s squee.

In a chronological feed with following and retweets people will form echo chambers. There are a few kinds of content that do well in that environment (cat pictures) but outrage, clickbait and such are at least 2/3 of it.

I’d imagine most people find Kiwi Farms somewhere between “not for me” and reprehensible but there are enough people who do want to be part of that community that it is a community. The argument that you just “shouldn’t go there” doesn’t hold water because the members of that community use it as a base to harass others.

To be specific, I have a model that can predict the comment/vote ratio for a story on Hacker News given the headline and things that are very high scoring are almost comically clickbait (right now in that feed I see “Panera Bread is phasing out its 'Charged Lemonade' drinks—why consuming too much caffeine can be dangerous”, “Why Apple’s ‘Crush’ ad is so misguided”, “Porn for women: Can it really be feminist or ethical?”)

An algorithmic feed can be like that or I could draw articles off the bottom of that list where you’d find papers about physics and molecular biology that might get 117 votes and 3 comments. (I made that model because I was posting a lot of those and people were telling me I should be posting things that people want to talk about…. I made the discovery that HN users really like to talk about cars)


Yeah but the people targeted by KF tend to deserve it. Their forum holds a significant amount of archived evidence of wrongdoing, deviancy and degeneracy. People ripping people off. Child abuse. Animal abuse. Sexual assault. It's almost always tied to actual things that normal people find worthy of criticism.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: