Words ‘created the internet.’ Now the Supreme Court may be coming for them
cnn.com
cnn.com
However, sadly, this isn't unexpected — governments all around the world are trying to restrict UGC or only allow it when the identity of the poster is verified. Can't let the serfs think out loud, lest it put the elite in jeopardy.
- The conservatives would be happy - The liberals would be unhappy
Conservatives' whole argument isn't "bad things I don't like on the internet" (that's the liberal argument), their argument is "the rules are rigged against my views".
The conservative position is free speech. The Twitter execs were just getting grilled in congress last week and all the conservative questions were "why were you rigging the rules against conservatives ie "trans ideology" written into the rules, emails where you suppressed content in accordance with FBI requests, etc
Meanwhile the liberals questions were: "why didn't you ban this conservative account? or this other conservative account? or this other one?".
Only 1 side here wants to ban accounts for their political views. The other just wants LESS restrictions on speech.
Remind me, again, who is banning books? Both parties in the US are awfully selective about when they will and won't support "free speech" as you seem to define it.
How about we focus on the actual issues rather than devolving to US team sport politics?
Edit: and incidentally the original comment was equally thought-terminating. But, why have a substantive discussion when you can go for easy polemics.
>Conservatives' whole argument isn't "bad things I don't like on the internet" (that's the liberal argument), their argument is "the rules are rigged against my views".
If the liberal rules are "no bad things on the internet" and the rules are rigged against conservative opinions.
What does that make conservative opinions?
The problem is exactly your comment: liberal opinions are not the same thing as the truth. So just because some liberal doesn't like your opinion (thinks its bad) doesn't mean you deserve to be deplatformed. There's no gotcha there with the logical endpoint.
Seems suspicious to me, I don't think right wing voices are actually getting silenced. The opposite in fact.
>In six out of seven countries — all but Germany — Tweets posted by accounts from the political right receive more algorithmic amplification than the political left when studied as a group.
https://blog.twitter.com/en_us/topics/company/2021/rml-polit...
https://www.theverge.com/2021/10/22/22740703/twitter-algorit...
So what does this all say about which ideas are more harmful? What does it say about who is actually getting silenced? The right wing media seems to love playing the victim.
Like all things, the pros and cons have to be balanced. Speech impacts people, whether turning them into modern Nazis or perverted fornicators.
Section 230 is predicated on the idea that sites like HN have no editorial function and therefore have no liability over posted content.
This safe harbor is extremely important and needs to be protected.
Unfortunately algorithmic recommendation makes that line very grey.
As a compromise I'd push for algorithmic transparency: show me how you're ranking content.
This would both clarify whether companies are engaging in deliberate editorial activities via algorithm (and therefore should be held liable for the content they promote) while also improving both user privacy (we'd be able to see how our data is being used) and online trust.
I honestly don't see a way to balance the two interests, unless a voluntary opt-out to the algorithm (such as HN's /new) also suffices to comply with the new law.
You just described how you think HN works but unless you work there you can't have first hand knowledge. I know I don't.
So show me the algorithm. Let's have an open conversation based on facts instead of speculation.
As an aside, I also think the lines are easier drawn than you think.
For example: is the algorithm based purely on overt signals from the userbase (votes, flags, etc)? Or is it based on an analysis of the content itself, which is then used to fuel a recommendation engine?
If the latter, are the recommendations based purely on analyzing past expressed preference? Or is there some other baked in weighting?
Algorithms aren't dark mysterious things. They're entirely explainable, and if opened up you could have a real discussion about what is and isn't editorial promotion of content.
Recommendations, sure, but not the removal of content.
What about my comments suggested otherwise?
There might be room for improvement into transparency of what is and isn’t allowed. But as of right now, platforms are allowed to be arbitrary. Publishing algorithms would make it easy to defeat spam filters and anti evil filters. They are black boxes for a reason.
No, it led to an internet where the listed things proliferate
And IMO they shouldn't be.
> Publishing algorithms would make it easy to defeat spam filters and anti evil filters. They are black boxes for a reason.
Security through obscurity.
Besides, if I have to choose between slightly poorer spam detection and algorithmic transparency on the sites that are setting the public discourse, I'm gonna pick transparency. Wouldn't you?
No. It’s my speech not his.
Let’s be real, the threat of benevolent censorship is what keeps this place civil and upstanding.
You might say, but but transparency. Now you are creating a new problem. For example, the Christian website above. If I go there extolling the virtues of Satan, you are forcing them to host copies of speech they find deplorable, in their public censorship reports. And if you obscure or roll up censorship reports, now you need independent speech removal auditors.
I would choose a better spam filter and a free market for you to make a competing service with transparency. The free market breaks down when institutions become proprietary and monolithic.
A better angle is the government to better force protocols, decentralization, federation etc. Even email is kinda controlled by two players, but at least if I got kicked off one I can still email all my old friends from my new account on the other. Data portability, service interoperability, and privacy better address a lack of transparency than forcing homogenous censorship rules on all sites.
I never advocated for any requirements regarding what can or can't be censored or promoted.
I am arguing for transparency, specifically because I recognize that codifying editorial standards could be exceedingly difficult.
Yet the vast majority of you comment is dedicated to standing up a straw man you erected.
Am I being confusing, or are you just shifting the argument to your preferred turf?
Opaque censorship should be an optional design choice private companies have the option of designing their products around.
I would expect any transparency requirement to be limited to only services with a userbase over a certain size, since the concern is with very large entities that have an oversized impact on public communication.
So if you're a small community or a startup, fine, no need for disclosure. But if you're a public company or have more than X users, I truly believe these companies have far too much control over public communication for us to be comfortable with the current model of total opacity in algorithmic behaviour.
That said, I could see a requirement for support commitments. Being able to get human support and transparent answers to account questions. If you can’t support people you accidentally lock out, you are too big. Like fire department requirements for a city population.
Another product I could see requiring regulation is login auth services. If I use my social media account to auth against other websites, you can’t just take that auth away from me if you don’t want to serve me anymore. Force a forwarder so I can use my new auth service to log into my old accounts.