the_last_pigeon
shiggy
it's look who it is
User ID: 62
Twitter dies for good in the next six months: 80% probability
By now you know that Elon gave staff a deadline of today (Thursday) to either commit to being "extremely hardcore" or leave (source). Unsurprisingly, most people - roughly 75%, according to some Internet rando - didn't take him up on this. Elon blinked and apparently people still have access.
That won't do much (WaPo):
“I know of six critical systems (like ‘serving tweets’ levels of critical) which no longer have any engineers,” a former employee said. "There is no longer even a skeleton crew manning the system. It will continue to coast until it runs into something, and then it will stop.”
But that's not even what I was going to write about, just what happened while I was composing the post. (Also let's put aside that he said "microservices are bloat" and then they killed the microservice serving SMS 2-factor login.)
To me, the biggest news is that he axed 80% of the 5500 contractors (source, Casey Newton, or someone with a premium account impersonating him I guess).
The contractors were responsible for things like moderation (source: what are they gonna do, use salaried employees?). If you don't have moderation for basic things like CSAM, you're boned. I know a thing or two about moderation, and if you let the Internet type into a text field, you get some dank shit. And crucially, you can't automate it away, because there's a human on the other side working to defeat whatever you're doing. I mean, the YouTube comment section probably has some of the most expensive automation on the planet working on it and the spam still gets worse every day, and I'm talking the obvious stuff like "HIT ME UP ON TELEGRAM <number>". The only thing that saves you is humans clicking buttons (and getting PTSD, but let's skip that for now). Google had 101k employees but 121k contractors as of March 2019, and that's what the contractors do, click buttons.
If you don't have moderation, you don't get the YouTube comments section, because they at least have contractors backed up by code (at the cost of many expensive engineer-years). You don't even get 4chan, because they at least have Those Who Do It For Free. You get some ungodly shithole most younger Internet users have never experienced. You're getting... the virtual equivalent of your local Greyhound terminal. Whatever happens to someone's chat room side project that gets posted to /b/. Sludge.
Twitter will have to either restrict posting to an unbearable degree or watch as the remaining users get tired of slurs in their replies and bounce.
Remember when Elon was just going to clean up the bots on Twitter?
(Reason for posting: I saw some takes elsewhere on this site that apparently Musk would lead Twitter to success or at least improve it or something, and disagreed.)
Wikipedia is deciding whether to discourage use of Fox News as a source in articles specifically for politics and science. As usual, please do not comment there unless you know your way around a Wikipedia discussion and can participate while following community standards.
In context: Wikipedia periodically holds discussions about the reliability of sources. It has a five-level ranking system for sources:
-
generally reliable
-
no consensus (= "we couldn't decide")
-
generally unreliable (= "usually don't use")
-
deprecated (= "never use")
-
blacklisted (= "never use", enforced in the wiki software)
The current discussion is about Fox News when it talks about the two topics of politics and science - for those topics, it is currently listed as "no consensus". For other topics, it is "generally reliable", and that status is not up for discussion here. Fox's talk shows are also listed separately as "deprecated" (= "never use"), and that status is also not up for discussion. There are 23 prior discussions listed about the reliability of Fox News for politics and science, starting in 2009 (although there may be more). This is the latest one.
Why this is relevant here: Wikipedia is a widely-used reference on the Internet (top ten websites globally, by number of visits) and Fox News is a well-known news source. The debate on whether Fox News is a reliable source for science and politics is thus likely to be of interest on this website.
Moving to the discussion itself: many points were raised of varying quality. There's quite a bit of back-and-forth and it's certainly not one-sided.
My take: while Fox is certainly useful for presenting facts that other sources don't, it's made factually incorrect claims that remain uncorrected. Those would make it difficult to use as the only source for a claim, and if you can't do that, what's the point. It can still be used for research while writing articles, like every other website on the Internet. As for the incorrect claims, various editors compiled lists of these; here's an 18-item list. I checked a few. Some were weak; some were worrying. For example, item 10 quotes from this Fox article: "PolitiFact appears to be shielding President Biden and Vice President Harris from criticism over their past rhetoric expressing distrust in the coronavirus vaccine during the Trump administration". Here's the PolitiFact page. It shows that "expressing distrust in the coronavirus vaccine during the Trump administration" is a misleading construction: Biden and Harris repeatedly emphasize that they would take a vaccine approved by public health professionals, but would not trust the sole word of Trump. Fox phrases it as during the administration, they expressed distrust in the vaccine, in general, but this is simply not what they did. Why that's bad: one could write a sentence in an article with that claim, and cite it to the Fox article, and that would be incorrect. The Fox article was published July 2021 and has not been corrected.
My take, part 2: The optics might not be great, but at least Fox still counts as reliable for everything but politics and science. I don't think they're managing the optics enough. Of course, it's a decentralized and anti-hierarchical community, so the odds they'd organically do something like that are low.
Where we go from here: Editors are requesting that the discussion be "closed" by a neutral third-party editor (or panel of such editors), and that may happen sometime soon. Editors are still adding comments to the main discussion in the meantime. The "close" can be appealed to the community, but if the closer does a decent job this is unlikely.
My credentials: I've edited Wikipedia for a while. I usually don't touch the politics side much.
While I can appreciate where you're coming from, by no means can the linked discussion be described as having a monolithic point of view. It would look very different if Wikipedia's community were as polarized as you suggest.
And the easiest way to fix any other issues you see is to do it yourself or convince someone to. I would like to make a standing offer to anyone reading this, that I will coach you through fixing any "political lean" problem of the sort you're alluding to. (Of course, a writeup should get posted here during/afterwards.)
Anyone can start one of these discussions about a source at any time, although in practice you have to gauge community mood at least a little. The key question here is whether you can pick a sentence from an article within the given topic and be confident that it's factually correct, and that's what's being questioned for Fox's politics articles.
If someone managed to put together a big list for, say, the NYT, I'd like to see it and I'm sure the community would like to see it. I agree without reservations that all major news sources should be subject to the same level of scrutiny.
For what it's worth: BuzzFeed currently has the same "no consensus" rating as Fox, although honestly on the strength of far less well-attended discussions.
For what it's worth, part 2: A discussion on MSNBC was launched after the last big Fox News discussion, but nobody put in the same amount of effort to find instances of inaccuracies (only one person posted, and they posted a mistake in a headline, and it is known that headlines aren't written by the article authors and are thus junk - see WP:HEADLINE). Thus the discussion reaffirmed that MSNBC is unusable for opinion pieces, as all opinion pieces are, and is generally reliable most other times, with the caveat that they don't even have written reporting on their news site so it's a bit of a strange discussion to have. (They have lots of blogs, though, which are all not suitable for use by policy.) If someone came up with a similar list for another major news outlet, I'd expect it to be taken seriously. I can't immediately find any examples of someone dropping a large list in an RfC on a "famous" left-wing source, but there's plenty to look through on the main "source reliability" list.
the_last_pigeon
shiggy
8mo ago
The contention is that he knew he lost, and despite that he did all of the above. I don't personally know whether that's true, but the prosecutor thinks he can prove it.
the_last_pigeon
shiggy
1yr ago
Not sure that Jan 6 impacted the Jan 5 runoffs too much, although I like the rest of your analysis.
the_last_pigeon
shiggy
1yr ago
Would you like to work together to improve that article? I would enjoy seeing a "bunch of admins" get in the way. Let's start a new post for it, though. I hope our overlords would be OK with that, as it is sort of culture-war.
Knowingly using sources that are lying is straightforwardly stupid and I'd hope that both (1) no actual Wikipedia policy can be construed that way, and (2) no discussion has concluded that way. What you linked is an unofficial and unsanctioned interpretation of policy. From the box at the top:
This is an essay.... This page is not... one of Wikipedia's policies or guidelines, as it has not been thoroughly vetted by the community. Some essays represent widespread norms; others only represent minority viewpoints.
I suppose that's on us, for bad signage. I'll certainly never argue that we're good at user interfaces :)
That said, a common criticism of Wikipedia is how it relies on existing sources rather than "the truth". This is an entirely valid criticism. It is a correct interpretation of policy that, as your essay says, Wikipedia would have advocated for a geocentric universe if it had existed back in the days when that was the mainstream viewpoint. In a sense, that's how it has to be. People fighting over truth itself doesn't make for a good encyclopedia, because verification of the results is many times harder.
the_last_pigeon
shiggy
1yr ago
I have yet to see evidence that rape is a failure mode of trans rights specifically and not just human nature generally. To restate, being trans does not cause you to be a rapist.
Generalizations aside, among all strategies to reduce prison rape, putting trans prisoners in with their birth sex population is surely one of the least effective ones. Anything else is honestly more likely to eliminate the problem entirely. For example, getting the guards to stop raping the inmates will make them more effective at preventing rape among the inmates, including rape done by trans people.
Housing sex criminals separately would be another good thing to try. In some of the articles and stuff I looked at while writing this post, some people proposed we house trans sex criminals separately. But housing all sex criminals separately would obviously address the problem too. Even if we had limited space, severity of crimes would obviously be a better way to prioritize than being trans.
I guess some people see other people getting worried about prison rape by trans people and think they're calling trans people rapists. Obviously they're not, but it's still fishy. To focus on prison rape by trans people instead (and, observably, it's very much instead) of prison rape generally implies there's something special about prison rape when trans people do it. Maybe one argument in favor of such specialness is that we can apply a very straightforward intervention (no trans women in women's prison) and it straightforwardly gets rid of this one subcase of prison rape. But there are even simpler interventions that obviously work better, such as not housing sex criminals with everyone else. In summary, it's more effective to target the "rapist" part of "trans rapist" rather than the "trans" part.
Can you substantiate that claim about Fetterman's mental health? I think it's more likely that his mental capacity will be fine by next January and it's just his speech and hearing that look bad.
See the full list of sources. Not only are all right-wing sources not listed fair game (even politically biased sources are explicitly allowed, see WP:PARTISAN), non-left-wing sources on that list listed as "generally reliable" include Reason, the WSJ, Deseret News, Financial Times, and Religion News Service. Non-left-wing sources listed as "no consensus", meaning they're usable based on context, include The Washington Times, The American Conservative, Washington Examiner, the Cato Institute, and National Review.
the_last_pigeon
shiggy
4mo ago
OK, so explain how overprovisioning and geosynchronicity were not correctly analyzed in the discussion. "It's wrong but I'm not telling you how" doesn't give me anything to work with.
the_last_pigeon
shiggy
4mo ago
That was covered in the discussion; overprovisioning and horizon/geosynchronicity introduce ~4x factors in opposite directions so the estimate is fine.
the_last_pigeon
shiggy
4mo ago
What does it ignore, overprovisioning, being over the horizon, space lasers, or other stuff? The first two are in the discussion, I don't know how you'd account for the third - yes, you got me, I'm not a domain expert - and I'd appreciate you expanding on if it's something else. Yes, I only took comments supporting my claim; nobody in the discussion was able to produce a satisfactory response to them, so although it's cherry-picking I'm not leaving out promising counter-arguments.
Carr's dissent, which the blockquote is from, is itself misleading in many ways. HN comment:
When applying for RDOF you say what service tier you are targeting and instead of shooting for the minimum 25/3, Starlink applied for 100/20. When they didn't reach those speeds[1], they were ineligible but not just because they didn't hit the required speeds on their existing network. There are more details here[2] but the jist is that Starlink bid to supply 100/20 internet to over half a million subscribers and the FCC was required to assess if Starlink was reasonably, technically capable of supplying those speeds by 2025. Starlink reportedly argued that once they can properly launch Starship, they can surely hit the required speeds. As of yet Starship hasn't had a successful launch. On top of this, the statistics that were available at the time showed that Starlink transfer speeds were already trending down and the network is a lot less utilized than it would be in 2025. There are technical challenges that need to be solved before Starlink is remotely capable of meeting that obligation and the challenges don't appear to be resolved yet. Giving Starlink money is a gamble and the FCC would rather play it safe.
RDOF rules set speeds of 25/3 Mbps as the minimum allowed for broadband service delivered by winners. However, participants were permitted to bid at four different performance tiers: 25/3 Mbps, 50/5 Mbps, 100/20 Mbps and 1 Gbps/500 Mbps. When the auction closed, the FCC noted 99.7% of locations were bid at 100/20 or higher, with 85% bid at the gigabit tier. That means Starlink will need to provide speeds of at least 100/20 in order to meet its obligations.
From subsequent discussion:
- the "decision today to revoke"
- "They decided a while ago (2022) that Starlink wouldn't get the RDOF grant. This was essentially an appeal to see if the decision would be reversed, and they upheld the original decision not to fund Starlink. It's not a check after deployment thing, it's a 'check if they actually can deploy in the first place' situation." (HN comment)
- "They didn't decide now. The program was created as a two step process initially. Starlink succeeded in the first round, but was denied in the second, more in depth, review that lead to the rejection. This was basically an appeal of that rejection. The second round was designed to eliminate providers who didn't seem able to deliver on their promises even with the subsidies. It was made to prevent a situation where either party (but mostly the US Gov and tax payers by extension) was on the hook for unsuccessful delivery." (HN comment)
- from the dissent, "The trouble with this argument is that SpaceX never indicated that it was relying on the Starship platform to meet its RDOF obligations, and in fact it repeatedly stated that it was not.":
- "Doing some math, currently each satellite launch sends up 22 satellites at around 2.8 Gbps per satellite. For each launch, Starlink adds [approx] 61.6 Gbps of capacity. If we cut that up into 100/20 slices, each launch supports 616 customers at 100/20. To support 650,000 subscribers at 100/20, it would take about 1055 perfect launches." (HN comment)
- "They need to do 180 a year [they've done 91 in 2023 so far] to put enough satellites up to even try to hit the 2025 deadline. That's not even counting any satellites which may fail between now and then and need replaced. This is a major reason why the FCC didn't think they could have met the 2025 obligation to reach [approx] 650,000 subscribers with 100/20 and rejected their application." (HN comment)
- "Doing some math, currently each satellite launch sends up 22 satellites at around 2.8 Gbps per satellite. For each launch, Starlink adds [approx] 61.6 Gbps of capacity. If we cut that up into 100/20 slices, each launch supports 616 customers at 100/20. To support 650,000 subscribers at 100/20, it would take about 1055 perfect launches." (HN comment)
tl;dr SpaceX's claim was not credible.
Actually I know some people who worked on Starlink. I may update this comment with what they have to say tomorrow.
the_last_pigeon
shiggy
1yr ago
Not a bug or suggestion, just being thankful - being able to view precise upvote/downvote numbers, and thread-view numbers, is a big quality-of-life improvement.
Other platforms won't feel the need to pay for Twitter traffic, because they don't depend on it. What this change will do is enrage users who find it useful or even financially necessary to link to other platforms.
OK, so what's your solution to Jews being overrepresented in these institutions, assuming you think it's a problem? Actually, why do you think it's a problem? I would make this comment more high-effort by guessing answers to those questions but I don't think I have a good enough mental model to be using it for that yet.
the_last_pigeon
shiggy
1yr ago
That discussion ended in your desired outcome with the contested sentence being removed. The current text of the article (can't bother to check how long it's been there, but this is the tool you would use for that):
In filings made with the SEC in March 2018, Gab stated that its target market is "conservative, libertarian, nationalists, and populist internet users around the world" "who are seeking alternative news media platforms like Breitbart.com, DrudgeReport.com, Infowars.com".
the_last_pigeon
shiggy
1yr ago
Right, there had already been a considerable amount of circus.
the_last_pigeon
shiggy
1yr ago
Desire to vote and political knowledge are spectrums and, I'd assume, correlated. Everyone above a certain level of desire to vote does so. Get-out-the-vote efforts lower the level. Thus I'd figure the resulting voters are not going to vote at random, because their level of political knowledge is better than that; you'd have to reach pretty amazing levels of turnout before you hit people with little enough knowledge to vote randomly or otherwise in a hideously misinformed fashion.
I appreciate the forecast and explanation. I disagree with removing votes, but only weakly. Votes reflect (among other things, and poorly) the effort level of a post, and thus showing higher-voted comments first improves the reading experience of the site. Votes aren't necessary for an echo chamber: ideological conformity can be enforced with social pressure (by getting lots of disagreeing replies). Without votes, we'd rely on mods more heavily to police low-effort comments. (Which isn't necessarily a bad thing, I suppose. Informally, most votes on the current CW thread look like the sort of votes you're trying to avoid.)
Other sites have various technical bodges that I hope might suffice instead of removing votes entirely. Hacker News and Lobste.rs require a fair amount of karma (500 for HN) to unlock downvotes, and Lobste.rs requires reasons to be selected from a menu for downvotes on posts. Furthermore, HN prevents you from downvoting direct replies to your comments or comments more than 24 hours old, and has a minimum score of -4 (beyond which downvotes don't change the score). Slashdot's M2 system offers a different route, perhaps one far too baroque for current Internet users. For an even wilder idea, maybe we could cap comment scores at +1/-1, and have all scores start at 0? Anyway, I think, but can't justify, that voting and tree structure go hand-in-hand, and I worry about removing one but not the other.
the_last_pigeon
shiggy
1yr ago
Would you not agree that "I'm going death con 3 On JEWISH PEOPLE" merits a stronger noun than "criticism"? Perhaps "hatred"?
the_last_pigeon
shiggy
1yr ago
Context matters tremendously when determining the reliability of sources, and their appropriate use on Wikipedia. Sources which are generally unreliable may still be useful in some situations. For example, even extremely low-quality sources, such as social media, may sometimes be used as self-published sources for routine information about the subjects themselves. Conversely, some otherwise high-quality sources may not be reliable for highly technical subjects that fall well outside their normal areas of expertise, and even very high-quality sources may occasionally make errors, or retract pieces they have published in their entirety. Even considering content published by a single source, some may represent high-quality professional journalism, while other content may be merely opinion pieces, which mainly represent the personal views of the author, and depend on the author's personal reliability as a source. Be especially careful with sponsored content, because while it is usually unreliable as a source, it is designed to appear otherwise.
Consider also the weight of the claims you are supporting, which should be evaluated alongside the reliability of the sources cited. Mundane, uncontroversial details have the lowest burden of proof, while information related to biomedicine and living persons have the highest.
Like you, I wish the average Wikipedia editor cared more about those rules.
Even the existence of the list itself is controversial in the community, precisely for the reason you articulate: that it tempts people towards a simpler system that does damage when applied in the real world.
The Vice President of the Confederate States of America (via Wikipedia):
Historian William C. Davis (Look Away pp. 97-98 via Wikipedia):
I won't tell you Northerners fought "to free slaves" because that's as obviously wrong to me as it is to you. It's easy to reconcile the seeming inconsistency in Mr. Lincoln's thinking: the South seceded to preserve slavery, and the North took up arms in reaction. Perhaps the North didn't care about slavery, but that doesn't matter in evaluating the South's cause.
More options
Context Copy link