yanderekko

The line isn't drawn at "literal Nazis", because for all the criticisms one can lob at KF it is clearly not dedicated to Naziism. It platforms Nazis, for sure, but if the argument is either that a large blast radius that hits a lot of non-Nazi content is acceptable or the old "well if you platform literal Nazis you are a literal Nazi, sorry" fallacy, it's not good.

I recall reading the other day about Germany making deadnaming a crime..

>My country has a censor's office, for example. Some 'speech' is literally illegal to possess.

And it's embarrassing. The KiwiFarms exchange with New Zealand after the Brenton Tarrant shooting (are you even allowed to mention his name? Hope I'm not getting you in trouble here) was one of its prouder moments.

The Constitution would not provide a carveout for the long-term censorship of a forum like KF based on some rationale like "creating a volatile atmosphere", so I don't know what you expect regulation to accomplish here. And of course activists would scream if such regulation were applied in a viewpoint-neutral fashion (eg. platforms being mandated to delete content that advocates for crime / violence - imagine how Trump would've used this power during the BLM riots!) - the non-neutrality is considered a feature of private censorship, not a bug.

Sure, there are 1A concerns there. But that's not part of the HE issue as far as I'm aware.

It's easy to have a principled stance against both net neutrality and private online censorship. Note that the EFF statement is not calling for legislation to stop HE's behavior.

Net neutrality stopped being a concern when activists decided that it was to their overall advantage to ally with powerful, censorious corporations in order to combat right-wing speech .

>I'd argue that there are legitimate safety concerns that trump speech in this particular case.

What makes HE well-positioned to be able to determine what constitutes a "legitimate safety concern"? What happens when a pro-life organization or a corrupt politician starts leaning on ISPs to enforce their own views of what constitutes a "legitimate safety concern"?

Cloudflare denied service to KF on an imminent threat to people's lives, which could have been justified on a short term basis. Their refusal to reinstate service months later strains the credulity of this rationale, and shows that it was never really the goal.

I'm going to guess that if we ever get a post-mortem on what really is happening inside HE, what we'll find is not "a well-reasoned, dispassionate debate occurred and a consensus was reached in favor of censorship". Instead it will be like "someone powerful in the company made a unilateral decision and dissent was absent because everyone understood that it would have negative career consequences."

I mean, to be fair this doesn't really pertain to the sort of Constitutional legal questions that the ACLU famously takes a hardline stance on. I don't think the ACLU endorsing private censorship is clearly at odds with their overall mission. (As opposed to say their going soft on hate speech laws, which clearly has 1A implications.)

At least payment providers can plausibly point to financial risks and legal concerns as why they engage in what comes off as moral policing. It's pretty clear that any similar claim on HE's part would be pretextual.

>It feels like private companies are de facto writing the laws about what's allowed online and in society right now, and that it's almost become a loophole for censoring free speech on a whim.

"Feels like"? This is clearly embraced as a necessary step to preserve democracy by a small-but-powerful segment of population.

I just don't believe that unless you have specific quotes you can point to that indicates this.

>A counterpoint — every day tons of financial disputes are handled by banks and insurance companies. Unlike with government, no violence is involved, and things go well. Why involve governments and violence when we can avoid the bad things without violence?

Good point, maybe the private financial institutions (or more!) of the country should lean into this and offer a shadow legal system whereby anyone can bring suit for any allowable reason with the ultimate threat being a sort of (nonviolent!) exile from digital society. That could surely not go wrong.

>Thankfully, the site is no longer available via clearweb

Untrue, but once enough people know this I'm sure it'll be taken out again. I'd post the link but I'd probably be jeopardizing my YC account..

The nexus between KF harassment and the suicides in two of the three cases is extremely tenuous ("KF had a thread X disliked" -> "X committed suicide" -> "KF induced X's suicide") and in the third the evidence that a suicide occurred to begin with is lacking.

I'm sure many, many more suicides could be linked more strongly to any of the major social media platforms.

In any case, the reasons for deplatforming have always been more diverse. CloudFlare dropped KF for representing some sort of imminent and ongoing threat to human life which you'd think a good-faith review of would indicate was misguided (and thus their service should be restored.) KF has not been able to establish relationships with other hosts not because literally everyone finds them repugnant, but because entering into a public relationship with any company immediately makes that company a target for a hate campaign, and even "free speech" hosts and such have found the cost-benefit math unfavorable.

No, even the more-panicked wing is not betting the proverbial house on AI killing us within the next few years, although the possibility is there. The "don't expect to live much longer" sentiment is based on a probability horizon that extends over decades.

>when the real risk is what people believe machine derived "facts" are

It's hard to play out a scenario where this leads to actual human extinction.

It's noteworthy that even if we do see "evil" it'll just be something painted on an underlying veneer of indifference, eg. ChaosGPT.

>So you can continue your performative concerns about Roko’s Basillisk or whatever topic of distant, paralyzing uncertainty is overflowing from the next room into mine.

Amusing, usually rationalists are made fun of because the concerns about Roko's Basilisk are not performative.

Going after mentions of "fields" is going to make 2024 a lot of fun.

>I'd loved to bring up old HN threads announcing the death of Facebook using similar broad strokes. Apparently Meta is the competent one now because they pigeonholed a cloned feature on their platform + the alternative is no longer cool among the tech crowd on a niche programmer/startup forum.

The intelligentsia hates Musk far more than Zuckerberg right now, and will cheer on anything that could potentially hurt him. There's also some wishful thinking that Threads will institute the sort of mass censorship of right-wing speech that was present on Twitter, but it seems unlikely that there will be very different standards than what you see on Facebook, which is often derided as a right-wing boomer-infested hellscape.

Someday, when the AI releases a nanobot swarm comes to kill us all, a philosopher's last words will be "yes, but is it truly intelligent?" before he is broken down into biodiesel that will be used to power the paperclip factories.

>If these people were so concerned, they'd by shouting from the hilltops and throwing their entire life savings into stopping us. They would organize workplace walkouts and strikes. There would be protests and banners. Burning data centers.

I think we underestimate the intoxicating lure of human complacency at our own peril. If I think there's a 90% chance that AI will kill me in the next 20 years, maybe I'd be doing this. Of course, there is a knowing perception that appearing too unhinged can be detrimental to the cause, eg. actually instigating terrorist attacks against AI research labs may backfire.

But if I only think there's a 25% chance? Ehh. My life will be less-stressed if I don't think about it too much, and just go on as normal. I'm going to die eventually anyways, if it's part of a singularity event then I imagine it will be quick and not too painful.

Of course, if the 25% estimate were accurate, then it's by far the most important policy issue of the current day.

Also of course there are collective action problems. If I think there's a 90% chance AI will kill me, then do I really think I can bring that down appreciably? Probably not. I could probably still have a bigger positive impact on my life expectancy by say dieting better. And let's see how good humans are at that..

Sure, but I wouldn't consider it reward hacking if we knowingly gave it the goal of maximizing papercips and it did this. Maybe I'm being pedantic, but that would be a pretty clear due diligence failure on the prompter's fault.

>Even if there was a way to gain efficiency through self modification rather than losing it, it would likely hit a plateau of diminishing returns, like most things in the world do.

And maybe you're fine with betting your life on that, but the people who are more-hesitant on this are not crazy.

Okay. When people say "the future is doomed", I assume they're talking about likely near-term scenarios and not the fact that eventually entropy will dissolve the universe or whatever.

> As always the answer isn't 'cool, let's regulate AI to limit its profitability, thus limiting AI development' but rather 'we should keep throwing money at it, just making sure we throw money at the right particular people.' Yudkowsky didn't want to bomb all data centers, just the ones that wouldn't comply with his regime.

Who would have to own the data centers for Yudkowsky to support "throwing money at profit-driven AI development"?

>And even if we achieve it, how do we know the AIs won't convince Homo sapiens to simply stop breeding and fade away in alignment with our values?

If it could do this, then we should seriously question whether it's actually a bad outcome.

Does it look so different though? What happens when the LLMs are good enough to propose practical code improvements that will result in more-useful LLMs? And when we start generating LLM-driven agents that actually train and deploy these more-useful LLMs? Are we taking on faith that we'll never get there?

This is what a lot of skeptics fail to appreciate imo: That yes the current generation of models are not threatening, and on their own they probably never will be. But there will likely come a point where they can be used to recursively self-improve, and that can potentially lead to very dangerous scenarios.

>These include reward hacking which is a huge problem with LLMs and why they lie and hallucinate so persistently, power seeking behaviour, and developing unanticipated harmful instrumental goals.

I'm not sure I'd call it reward hacking when everyone understands that LLMs are just trying to engage in text prediction and that this can often be a poor proxy for trying to generate "useful" text. A paperclip maximizer that ends up turning the world into paperclips rather than creating a new cool paperclip factory has an alignment problem, but it did not reward hack.

>As it stands now, the future without AI seems pretty doomed too.

As pessimistic as the current zeitgeist is, very few imagined futures reasonably result in actual human extinction.