Hacker News new | past | comments | ask | show | jobs | submit login
Only distributed fact-checking can keep up with democratized distribution (abe-winter.github.io)
132 points by awinter-py 10 months ago | hide | past | favorite | 182 comments

I agree with the author that optimizing for virality as the bottom line is a disaster than can only go on for so long. Its consequences are obvious at this point, and it's not leading to good places anywhere it is applied en masse. (For those unconvinced, I encourage to read about how the most viral and captivating content is often content that purposefully instills anger, fear, outgroup-hatred, schadenfreude, etc)

But deciding that we 'just need to fact-check' things more seems to be a very incomplete solution. Much is said about "who gets to fact-check?" but I think even if we assume the people fact-checking are good (morally) and mostly very correct (veridically), this solution will still fall extremely short.

Even if you perform good fact-checking, there is little evidence that this will achieve your goal, which is not actually correcting text on the Internet, but correcting peoples’ beliefs in their own minds, which turns out to be pretty difficult (telling people they are wrong while removing content they want to read is likely not the best way, even so).

Remember that many rights that you take for granted were only gained thanks to the failures of past powers to control the narrative as strictly as they wanted to. While I don't think that anyone can control 'the narrative' so to speak, especially not a single company or government (that currently exists), we will still see ceaseless attempts at such; the temptation to correct everyone is just too great (https://xkcd.com/386/, anyone?), especially if you believe you're on the side of Virtue (which it turns out, almost everyone believes).

I have been toying with the idea to create a Wikipedia Cloning Platform that would allow people to basically host their worldview on a subdomain, so they can change and edit wikipedia along their perceived worldview without external edits, and even import "truths" from other fringe worldviews, while maintaining links to the base Wikipedia where they see fit.

At first this sounds like the worst idea ever as it would allow even more bifurcation of the global consensus narrative, but I believe having all this in a structured manner would allow for a better discussability of the topics in question and allow for a place to store arguments on topics in a structured long form instead of asinine twitter discussions or spats blogs or comment sections of newspapers.

Lawful content moderation while trying to keep an unbiased reputation would probably be a nightmare though.

I think that would be great, though I've been thinking of problems that would make it likely not to work that well.

There was one fork called "Conservapedia" which aimed to be a conservative POV version of Wikipedia. But they did a bad job in terms of quality and engagement, and it's moved more and more in the direction of political advocacy until all that anyone edits or that appears there is political advocacy content. You could contrast this with the old Catholic Encyclopedia (https://www.newadvent.org/cathen/) which is more like an actual encyclopedia that someone, including a non-Catholic, would actually want to use to get the Catholic perspective on a topic.

Wikipedia's NPOV policy is impressive and the community's efforts to apply it are impressive, but it's ultimately very difficult to find an objective NPOV account of things, and one that's satisfactory to everyone. One difficulty here is the Undue Weight policy (which, to be clear, is probably necessary to make NPOV workable). This says, in effect, that it's considered NPOV to mention something at greater length than its relevance to the topic at hand warrants, or in a context where it's not particularly important. Unfortunately judgments of relevance, salience, or importance are themselves quite POV and Wikipedia's solution is mostly to rely on "reliable sources" -- typically published journalistic and academic writing. But journalistic and academic writing itself contains all sorts of biases (many of which Wikipedians are extremely aware of and have discussed at some length), including the likelihood that nobody will have actually written a journalistic analysis of the Mormon/vegan/Russian government/left-anarchist/Hindu nationalist/Reformed Christian/Maoist/gender-critical/Extinction Rebellion take on specific person or topic X. Which could very likely, if it existed, place much more weight on some aspect or issue than would be considered "due" in most other published analyses.

You can also read many Wikipedia articles and see that, to the extent they relied on sources, the sources all held the view that the subject of the article was good/legitimate/appropriate/normal, or was bad/illegitimate/sketchy/questionable, and it is in principle "NPOV" to not report on any other views (if they don't have reliable sources mentioning them). However there must be people who have substantive reasons to take another perspective but don't have the ability, resources, motivation, or desire to produce reliable sources reflecting that perspective. In the absence of that, those views can be described in their own article but can't be used to color any other article.

I think forks always have a huge network effects problem. There was a famous Wikipedia fork called Enciclopedia Libre y Universal en Español, which started as a result of ideological disagreements with or concerns about Wikipedia.


The Enciclopedia Libre was bigger than Wikipedia for two years but ended up getting surpassed by Wikipedia, I think largely due to network effects.

Even now if you wanted to make a fork based on an ideological disagreement, or in order to be consciously and avowedly POV in some way, you would have to find people who wanted to actively edit and maintain a whole encyclopedia with high quality standards -- which is probably tough because Wikipedia has, by default, the mindshare, brand, infrastructure, and community to be the most comprehensive and up-to-date.

I don't see an obvious way out of this quandary, even though I would love for there to be more thorough, high-quality POV encyclopedias.

The idea is to not address this quandary at all, but to let individual people organize their own "Worldviewpedia" according to their belief. I would have roritharr.worldviewpedia.org and could start writing my drivel down there, and maybe copy something I agree with from schoen.worldviewpedia.org and some other stuff from infowars.worldviewpedia.org.

This would allow me to fully rationally work through those topics and connect them to adjacent topics of the global consensus narrative I still believe in.

This may push some people into an ever deeper rabbithole, but it might also cure a lot of the "5 min expert" syndrome we've seen.

>This may push some people into an ever deeper rabbithole, but it might also cure a lot of the "5 min expert" syndrome we've seen.

But people being pushed into an ever deeper rabbithole of lies and extremism is a much bigger problem than "5 min expert" syndrome, and the former inevitably feeds the latter.

Sure, people have the right to their recursively self-reinforcing reality bubbles because freedom of speech and all, but societies decohere when their members violently disagree on even the fundamental aspects of reality.

The decoherence is in some sense inevitable but the proliferation of viewpoints will make it to some extent self-correcting and prevent any particular viewpoint from gaining hegemony without being grounded in pragmatic utility (which is not to say that it will be ground in truth).

Wouldn't maintaining a fork of Wikipedia all on your own be a huge about of work with relatively little benefit in terms of helping other people learn about your views and your knowledge?

It would be a huge amount of effort just to keep up with other people's mainline Wikipedia edits on a handful of articles that you already knew were relevant to your interests, and prospective readers would have a super-hard time finding your versions, and also distinguishing "Wikipedia articles that you had the opportunity to edit and keep up with" from "copies of mainline Wikipedia articles that you actually disagree with, but that happen to be present in your fork because you haven't had the time to edit them, or you haven't even noticed that they're there".

Or would you have a fork in which the only articles actually present were ones that you had edited, and all others produced a redirect to Wikipedia with an interstitial message indicating that you hadn't had any opportunity to review that content?

>Or would you have a fork in which the only articles actually present were ones that you had edited, and all others produced a redirect to Wikipedia with an interstitial message indicating that you hadn't had any opportunity to review that content?

Something along those lines, the work is kind of the point. Maintaining a coherent but alternative worldview on your own is a gruesome task, the more neatly it fits into the global consensus, the easier it is going to be to maintain, allowing for radical ideas which are nonetheless consistent with f.e. our laws of physics would be much easier than trying connect free energy to the laws of thermodynamics.

Any tech solution to this problem gurantees that it will make the human problem worse.

I guarantee this.

Any human behavior first approach, will have some marginally better but equally troubling progress.

This is a hell problem.

actively edit and maintain a whole encyclopedia

What about just maintaining a set of annotations (highlighted passages, sidebars, extra references, etc)? Wikipedia would remain Wikipedia, and your overlay-pedia (?) would be a wrapper. Major edits to a heavily annotated page could trigger a notification to someone who could then try to update the annotations.

I would use this if it was open source and self-hosted.

> Wikipedia's NPOV policy is impressive and the community's efforts to apply it are impressive, but it's ultimately very difficult to find an objective NPOV account of things, and one that's satisfactory to everyone. One difficulty here is the Undue Weight policy (which, to be clear, is probably necessary to make NPOV workable). This says, in effect, that it's considered NPOV to mention something at greater length than its relevance to the topic at hand warrants, or in a context where it's not particularly important. Unfortunately judgments of relevance, salience, or importance are themselves quite POV and Wikipedia's solution is mostly to rely on "reliable sources" -- typically published journalistic and academic writing. But journalistic and academic writing itself contains all sorts of biases (many of which Wikipedians are extremely aware of and have discussed at some length), including the likelihood that nobody will have actually written a journalistic analysis of the Mormon/vegan/Russian government/left-anarchist/Hindu nationalist/Reformed Christian/Maoist/gender-critical/Extinction Rebellion take on specific person or topic X. Which could very likely, if it existed, place much more weight on some aspect or issue than would be considered "due" in most other published analyses.

> You can also read many Wikipedia articles and see that, to the extent they relied on sources, the sources all held the view that the subject of the article was good/legitimate/appropriate/normal, or was bad/illegitimate/sketchy/questionable, and it is in principle "NPOV" to not report on any other views (if they don't have reliable sources mentioning them). However there must be people who have substantive reasons to take another perspective but don't have the ability, resources, motivation, or desire to produce reliable sources reflecting that perspective. In the absence of that, those views can be described in their own article but can't be used to color any other article.

This is an excellent summary of the issue with wikipedia and how the rules that were set up to eliminate biases are unable to deal with certain types of biases.

I think its possible to improve on this situation (I'm not sure how but I think its possible) but to some extent the problem is unsolvable because there are no bare facts absent interpretation. Ultimately you're just going to repeat information you heard elsewhere and the decision about which information to include and which to exclude is inherently value-laden (i.e. biased).

> allow for a better discussability of the topics

or give nut jobs more validity than they deserve. it used to be one guy in the corner of a bar spouting off his mouth and that was his sphere of influence... now that guy has a megaphone and the smarter people are mostly doing things and don't have the time or the inclination to say "no that's obvious bullshit because x y z" so the crap rises to the top.

In most respects, this was distributed fact-checking: the 20 people who heard him, most of them went "that guys crazy" and relayed "I saw a crazy guy at the bar" rather then took a tape-recorder and then played back exactly what they heard to their social group of 20 people (and then repeated it endlessly).

Agree. No guarantees, but I think the most reliable way to expose bad ideas is to actually expose them. Shine the light of critical examination on them. Don't try to stuff them into a closet and lock the door.

This idea needs to be put to rest. Shine the critical light of examination on this "belief" that the most reliable way to expose bad ideas is to expose them.

Have you seen this tested? Have you evidence that this works?

Crucially, does this work at internet scale or in social medai-esque content environments?


I can say that we know for a fact, that exposing conspiracy theories by letting people broadcast them independently is a sure shot way of having them infect more people.

We've seen this evidenced all over web forums, and social media. Bad ideas get amplified. Worse, the more complex the logical counter argument, the far better the bad idea will fare.

Here are what we know, for certainty:

1) People DO NOT want to spend cognitive power.

1.1) Content that forces cognitive expenditure will gain fewer views and therefore be harder to go viral.

2) Emotion driven processing is cheaper.

2.1) Therefore emotion engaging content will proliferate

3) this is worsened over general demographics

3.1) Proofs or logical arguments that require more than common knowledge will need novel approaches to propagate.

Our human web of knowledge is a growth medium that suits conspiracy theories, self referential beliefs, and superstition.

Reason is the interloper here.

This is not even remotely what I was saying.

From the article:

"I wonder if there’s a similar effect in the antivax movement – autism seems to really be on the rise (0.7 to 1.5% in this century)."

CDC is reporting that Autism has increased MUCH MUCH more than this. By their own reporting standards it was 1 in 150 in 2000, to 1 in 68 by 2010 (https://www.cdc.gov/ncbddd/autism/images/autism_data_graphic...) and the exponential increase has continued.

The idea that Autism and Vaccines is not linked is fake news. The CDC's own experts have blown the whistle saying their research was faked at the behest of the CDC to engage in a coverup.

An entire book was released on this subject: https://www.amazon.com/Vaccine-Whistleblower-Exposing-Autism...

And the semi-secret vaccine court which is heavily stacked against vaccine injury claimants, recently found that Autism was caused by a vaccine and paid out millions in damages. https://www.cbsnews.com/news/family-to-receive-15m-plus-in-f...

There are secret studies buried by the CDC showing profound increases in auto-immune diseases that RFK Jr. had to dig up via FOIA, example: https://twitter.com/Perpetualmaniac/status/12040577205858836...

The idea that anti-vaxxers are misguided and need to be algorithmically silenced by some form techno-fascism will be seen as the scariest time in human history. This is nothing but a coverup for big pharma. It will go down in history as the biggest medical coverup in human history, dwarfing the "cigarettes are healthy for you" that the CDC pushed in the '60s.

The fact that the left is pushing this anti-free speech mantra is absolutely disgusting. The left used to stand for liberal values like free speech. Now they stand for big pharma, big tech and chinese style censorship to invoke cancel culture.

How shameful!

> "I wonder if there’s a similar effect in the antivax movement – autism seems to really be on the rise (0.7 to 1.5% in this century)."

> CDC is reporting that Autism has increased MUCH MUCH more than this. By their own reporting standards it was 1 in 150 in 2000, to 1 in 68 by 2010 (https://www.cdc.gov/ncbddd/autism/images/autism_data_graphic...) and the exponential increase has continued.

What do you think 1/150 and 1/68 translates to? now ask yourself if you are that wrong about a fundamental condition of your argument what else could you be mistaken on?

> What do you think 1/150 and 1/68 translates to?

It translates into the increase in ONE DECADE. Now go read the claim by the author: that this increase happened in ONE CENTURY.

You do realize that this century started only 20 years ago?

Even in this interpretation (you are being very generous) he's off by a huge margin. The the rise of autism follows an exponential curve.

I don't think I'm being generous, "this century" unambiguously means the current century starting from year 1 (19th, 20th, 21st) not the previous 100 years. See for example the usage in https://en.wikipedia.org/wiki/21st_century

One example of what you're talking about is the conservapedia: https://conservapedia.com/Main_Page. Although it doesn't allow tracking edits like you mentioned.

In another post I pointed out that I think Conservapedia is a bit of a tragedy. An encyclopedia, with or without an explicit point of view, should ideally be an interesting and useful reference work. That goal needs to be front-and-center, over and above questions of how to resolve editorial disputes or whatever.

An encyclopedia with a point of view should still have high quality standards and a focus on collecting information that readers will want. Conservapedia seemed to instantly go in a direction of "hey, a forum to score political points! whoo!!".

One thing that might help (apart from cultural issues about people's motivations for contributing, maybe?) would be to say that articles are still expected to present all significant points of view (just as Wikipedia NPOV aspires to do), without the constraint that the article has to avoid favoring one of them (or has to give them weight in proportion to their presence in particular media outlets). I've seen the Catholic Encyclopedia and Jewish Encyclopedia try to fairly explain other religions' histories, practices, and tenets, and sometimes then mention the ways that those religious contradict Catholic or Jewish teachings. The main intended use of this information may be missionary or counter-missionary, but it's trying to be careful and thorough and also not tendentious -- it's not like "we have to slip in a THIS IS TOTALLY WRONG AND DUMB AND SO ARE PEOPLE WHO BELIEVE IT remark every other sentence when writing about other belief systems".

Also, trying to edit and add encyclopedic content that isn't just focused on disagreements with other points of view or positions...

One thing I was just reminded of is Paul Graham's Hierarchy of Disagreement:


A great thing would be for consciously-POV encyclopedias to always aim as high up on the Hierarchy of Disagreement as they can, on all topics. Imagine having a POV encyclopedia article on the Israel-Palestine conflict, or the U.S. abortion debate, for example, that argued or concluded that one side was "right" but still tried hard to fairly describe the beliefs and motivations of the "wrong" side and its supporters, and also mentioned historical events or facts that reflected well on the "wrong" side or poorly on the "right" side. More generally, one such that reading it made you genuinely more informed about the background and context of the issue, and better-equipped to have meaningful, substantive conversations about it with someone with a different background or opinion!

I agree on a basic level...some sort of a hierarchical system, a compartmentalized partial context system is definitely in need. The “full truth” is often too much for an average joe to handle.

I’ve seen “so that’s what those corrupt reporters from that whatever Lancet magazine want us to believe” wrt COVID...you can start from anywhere but no chance loading necessary contexts in full into a wonders of the world like that.

I don't know if this will work or not, definitely worth experimentation. This is a true original idea. Hearing for the first time.

> the most viral and captivating content is often content that purposefully instills anger, fear, outgroup-hatred, schadenfreude, etc

Another aspect is that on social media, virality isn't just based on people actively sharing stuff, but also content being so captivating that the platform amplifies it into virality based on user interaction statistics. (E.g. high click-through rates being used as a feedback signal to show content to more people, thus amplifying clickbait.)

I experienced that first-hand with a YouTube channel I started recently: one day I noticed the impression count (i.e. the number of times YouTube showed a video thumbnail to someone) growing exponentially. But view counts didn't grow nearly as much. So my content didn't suddenly get more popular, it had just randomly gotten a boost in the recommendation algorithm, showing it to many more people, most of whom weren't interested. (The boost subsequently disappeared.)

I think if social media were just people talking to each other, we'd be mostly fine. But imagine how little time you'd spend e.g. on YouTube if the only way to watch a video were to have someone send you the link. Obviously not enough for the platform.

So the search for virality turns into a self-fulfilling prophecy: if the recommendation algorithm thinks something will go viral, it can make it happen simply by pushing it in front of enough eyeballs.

Very few people are really interested in epistemology: https://jakeseliger.com/2020/11/10/personal-epistemology-fre...


These outright falsehoods that people take for truths are not taken for truth in a vacuum. As cliché as this may sound, people want to believe these things. So they believe them.

I was reading somewhere that people tend to readily assimilate information conforming to their own world view, and have a tendency to block out information that does not conform to their world view. It's claimed that these tendencies pre-date the internet.

If that's true, the negative effects of virality were inevitable.

Tribalism is as old as humans, and previous successful attempts at “defeating” it have typically involved warring tribes being wholesale eradicated by, or assimilated into, larger tribes, then states, then blocs. Now we have global tribes and nobody bigger left to defeat or assimilate them.

The only way out I see is to try to grow the “sensible majority” into a dominant majority that can robustly eject extremists - make a bigger tribe, essentially. But what cluster of issues unites enough people to form a dominant political force, while also rejecting the dark patterns that consume the current political tribes?

> The only way out I see is to try to grow the “sensible majority” into a dominant majority that can robustly eject extremists - make a bigger tribe, essentially.

That's only good if we have the ultimate truth - if we have no entrenched wrong ideas ourselves. I think that's a dubious proposition. If we have an entrenched false idea X, then anyone who denies X is one of those extremists that we eject. But if X is actually false, by ejecting them we destroy the ability of society to figure out that X is in fact false.

I think it’s the entrenchment that’s the problem. Science, for example, has a pretty good track record of coping with false ideas by being self-correcting and avoiding dogmatic truth claims. I don’t think a political body needs to be in possession of ultimate truth to identify and disempower those who do not seek it.

I’d guess economics is the only thing that can really unite people across all cultures. I live in an upper middle class Midwestern suburb , and my neighbors are Indian, Chinese, Vietnamese, Jordanian, Nigerian, Kenyan, and the people we bought our house from had Nepalese names based on a few Google searches.

If we just make everyone rich, maybe they won’t want the world to burn?

Having disposable income is a culture itself, and different levels of wealth/income can be broken down into different tribes as well.

Economics is also at the root of tribal disagreements. It’s all about securing assets for your tribe, and that doesn’t just mean money. It could also be the unquantifiable qualities that make up a stable tribe, which are found less frequently in tribes with less security of housing, food, energy, and opportunity.

“You are the company you keep” is a popular saying that sort of touches on it.

So, like... the Democratic Party? I'm less than half joking.

At least in the US (and oversimplifying) "tribalism" is what keeps the conservative party intact, while the liberal party is a big tent. One way you see this is that liberal policies tend to be broadly popular when presented outside the partisan frame. Meanwhile Republicans eneltered the election without even an official policy platform.

This is also ignoring the other dimension of lying: it's a way to assert dominance. The lie is the point.

You saw this when Russia invaded Ukraine: Vladimir Putin and other Russian propaganda efforts spout easily disprovable lies about what's happening, about what was plainly visible in the news (Russian soliders and vehicles crossing the border).

Russia wasn't trying to hide the lie, the point of the lie was to make you waste time fact-checking it, and then do the thing they needed anyway - which was you had to repeat the lie and then expend disproportionate rhetorical resources disputing it. And then they can simply turn around and tell another lie.

The idea that better fact-checking somehow will solve any problem is ignoring what the point of modern propaganda is - it's about dominance and power.

This should be upvoted more, very well said.

I usually frame it in terms of incentives but yes, you are absolutely correct, there is a power play involved.

> As cliché as this may sound, people want to believe these things. So they believe them.

Right, and there we also see the answer. We need to move people to no longer want to believe the lies.

For example, if I believe that it's all a conspiracy by "them", then it's not my fault that I'm a failure. That's comforting. But it also follows that I'm helpless, and that things are never going to get better. That's... not so comforting. If someone shows me that, I may not want to believe that any more.

Now, that's a very slow process, and almost requires one-on-one interaction with lots of time. You can't do it in a drive-by post on the internet. Worse, even if you put in the time and effort, it still may not work. But I don't know of anything else that will, and this might work at least some of the time.

The funny thing is the author acknowledges what a real solution would be - distributed fact checking, in which we can't say one institution or panel of experts has a monopoly on arbitration of the Truth™. If it were done in a way that made the consumer feel as though they were a part of the process, they might be more to allow some moderation even if it goes against their prescribed beliefs. I think that's what wikipedia got right.

The deeper issue that the author didn't really address well is how mis-incentivized online engagement is and the opportunities that exist on realigning those incentives. They discussed the marketplace of ideas and linked to how Facebook would lose engagement if they cracked down though, so I can't be too hard on the author.

My own solution is newsbetting. Place a wager on the political bias of the source of a de-titled article. Or a similar extension to whether a fact agrees with some established scientific entity. It puts skin in the game and allows consumers to feel engaged.

Totally agree. The other dimension that folks often overlook is the choice of what to fact-check. Droves of fact checkers that only ever check one side of an argument still present a bias.

> (For those unconvinced, I encourage to read about how the most viral and captivating content is often content that purposefully instills anger, fear, outgroup-hatred, schadenfreude, etc)

I call this "outrage porn" (not from me).

I like the idea, but the author doesn't seem to care enough for the root cause of dissinformation. That is, at it's core, the existence of conflicting interests. Because if literally everyone was on the same page, there wouldn't be a reason to lie about anything (of importance), right? Although distributed fact checking sounds like a good idea (however problematic it could be, even in a world without interest conflicts) how does the author prepare for the inevitable moment when a lie will get the seal of a Fact™ because big players push towards that direction? How would you protect your ecosystem against that? You are back to square 1.

That's essentially what already happened to the whole "fact checking site" industry. When journalists committed a mass professional suicide in their desire to help politicians they like, there was a perceivable need for non-propaganda information. Thus, "fact checking" gets into fashion. Big players immediately move in anf fact checkers start first getting subtle bias and then just do the same the press did only a short while before - declare that their mission are "bigger than just checking facts", they should serve the noble causes and help people make the right choices - those choices being their preferred partisan points - and surely attacking your own party with the same vigor as you would the opposing one is not helping people making the right choices... and in a short while we're back to the proverbial sq.1 I don't know how to solve this - I have observed several project that started with a noble goal of delivering just facts, away from propaganda and partisan bias - and in a year or two they become as biased as any MSM they wanted to replace.

That is very interesting. Can you provide a couple resources?

Snopes? It has been pretty good and neutral resource when they did urban legends. Once they got into big politics, they became very biased and then made a laughingstock of themselves "fact checking" labeled satire.

There's an angle here of [conflict theory vs mistake theory](https://slatestarcodex.com/2018/01/24/conflict-vs-mistake/). Your post seems to take a conflict-theory only perspective - that disinformation is caused only by conflicting interests, and never by incorrect analysis. Is that your opinion or is it enough to say there are some conflict theorists, therefore fact checking can't solve all the problems?

Interesting article. I think black and white is never a good rule, so although I would side on the conflict-theory perspective it would be very naive of me to claim that incorrect analysis doesn't ever contribute to dissinformation. Same goes with many of the points in the article that the author assigns to "mistake-theorists" but I can see myself subscribing to these.

With those in mind, I'd say that although not the sole cause of dissinformation, conflict of interests is the most important and impactful factor.

Everyone knows different beliefs happen because of mistakes, conflicted interests, and that differing models can still embody one truth (E.g., Newtonian physics vs relativistic), with different tradeoffs. Also, mistakes happen in even recognizing your own interests. A lot of mistakes are also practically unsolvable. Any moderation algorithm that assumes conflict away is hopelessly naive.

Who arbitrates?

Is it whoever owns big corporations? Is it the mainstream media? Are they trustworthy? Is it the government? Scientists? Which one of us holds the truth?

One man's facts are another man's lies. And this is fine. We should be able to tolerate multiple opinions. Labelling things as 'hate speech' and denying people the right to speech doesn't help!

I think C S Lewis said it best:

"those who torment us for our own good will torment us without end for they do so with the approval of their own conscience"

Why not jump off a bridge? After all, Newton's theory of gravitation is just a theory, and besides Einstein proved it's incomplete. The elite still pretends their disproven theory is truth because they want to control your life. Show the power of positive thinking and faith, make the leap!

But seriously, equating fact-checking with torment is probably not what C. S. Lewis had in mind. It has historically been the case that when opinions and ideology replace facts, it opens the door to tormentors who won't hesitate to hurt you to impose their version of reality.

Here is the full quote. I think its even better:

"Of all tyrannies, a tyranny sincerely exercised for the good of its victims may be the most oppressive. It would be better to live under robber barons than under omnipotent moral busybodies. The robber baron's cruelty may sometimes sleep, his cupidity may at some point be satiated; but those who torment us for our own good will torment us without end for they do so with the approval of their own conscience. They may be more likely to go to Heaven yet at the same time likelier to make a Hell of earth. This very kindness stings with intolerable insult. To be "cured" against one's will and cured of states which we may not regard as disease is to be put on a level of those who have not yet reached the age of reason or those who never will; to be classed with infants, imbeciles, and domestic animals."

Lewis's more philosophical works are massively underappreciated for how insightful, nuanced, and prescient they often were. If I were the kind of omnipotent moral busy-body he warns about, I'd swap every copy of 1984 in every primary school with a collection of his essays.

This presumes that the robber baron has a conscience which perceives the suffering of an outgroup to be a bad thing.

No, it explicitly doesn't. The implication is that the robber barons are pure self-interested evil. Self-interest means they aren't focused on you, as opposed to well intentioned self appointed saviors.

Is fact-checking tyranny? No.

If the only information allowed is the one approved by designated "fact checkers" (the ideal most major social networks strive for, and most fact-checking enthusiasts would wholeheartedly approve) - it is the most complete example of tyranny (at least in the information space) possible.

Of course, if fact checking merely exists as a set of opinions - it doesn't become tyrannical by itself. It's just an opinion of somebody on the question of what the facts are. But we all know it's not how it works. At least not how it works on Facebook, Twitter, YouTube, etc.

People who wish it to be difficult for millions or billions to read anything they deem "false" is accurately described as "a tyranny sincerely exercised for the good of its victims".

It also has been the case that when somebody is going to subjugate you, one of the first things they'd try to do is to control your information intake. If they can define what is "fact" for you, the rest goes much easier - after all, if somebody resists accepting "facts", it is ok to force them to do what "factually" needs to be done, it's only for their own good - it's literally like saving somebody from jumping off a bridge. So if you put a couple of bruises on him or break an arm or a leg, it's still a good deed - after all, he was going to jump off a bridge, so he's better off alive, even with broken leg. As soon as you've given yourself an exclusive license to facts, violence against and subjugation of whoever disagrees with you happens very easily and very naturally.

The difference is that Newton's theory of gravity comports with what every single person experiences in real life, in the same way. So nobody's really prone to believe it's some sort of hoax.

If we were all floating in zero gee in a giant space ship there would definitely be people arguing that it’s not real.

I wonder if flat Earthers seriously argue against Newton’s theory of gravity?

They may not argue against the idea itself, but the flat Earth just doesn't make sense if you account for gravity.

I don't know what flat earthers believe about gravity, but they do experience it the same as everyone else. They can drop an apple from height X and time how long it takes to hit the ground and come to the same conclusions as anyone else at least as far as the math is concerned.

> One man's facts are another man's lies

This is pretty absurd. We should be fighting disinformation instead of legitimizing it. Not all “facts” are based in a healthy reality. See: 2020 election “fraud”

This comment has been hidden because it was fact checked and determined to be partially false.

Reason: facts don't have to be based in a "healthy reality" to be true since perception of reality is subjective.

Click [here] to see original comment.


This looks pretty factual to me. It's fraud of some kind.

See also: 2016 Russian collusion.

The Mueller investigation indicted 34 people.

Excellent example of the dangers of "fact checking", where highly partisan talking point about "collusion", supported by a highly partisan, highly groupthink-selected source is presented as "fact". And supported by another statement which is a fact, but doesn't equal to the original one. Now it's your choice either to say "here you see fact-checkers confirmed/debunked it!" or spend infinite amount of time on fact-checking the fact-checkers, thus making the whole enterprise net negative.

So, it's prosecutors that decide what is true? What if they are wrong?

(I'm not talking about this particular case)

I have an idea, we can convene special committees where both prosecutors and defendants are given a structured format in which they can present all the evidence for guilt/innocence and counterevidence and arguments. We'll have separate professionals in charge of making sure the structure/rules of the committee are followed so that it's fair, and of course we'll have advocates well-versed in these rules on both sides, but maybe we'll also have a panel of everyday people representing, well, everyone, so that decisions like this remain a public matter as well as professional opinion.

Think of it as something like the royal courts for important decisions of old. In fact, that might be a good name for it: court.

Except courts do not determine truth. They determine legal guilt or innocence. They are actually designed this way - everybody can know somebody murdered somebody else, but if the court procedure is not followed and proof is not presented in a certain manner to achieve certain legal outcome - no prosecution. Moreover, there are special rules when known true evidence can be dismissed for not following the rules. OTOH, there are plenty of factually untrue court decisions - people have been incarcerated and even executed for crimes they factually did not commit. Check out Project Innocence, for example. If you want to find truth, courts are not the best place to look for it.

> Except courts do not determine truth

Courts definitely rule on matters of fact. Or truth. The fact that they also have a responsibility to apply the law once matters of fact are established doesn't exclude the fact-related portion of their focus and responsibility. Where known true evidence is dismissed, it's usually to protect defendants against violations of due process or other important rights.

If your point is that courts have limits in their ability to bring truth to light and apply it... sure. Every human being and every institution has such limits. If your point is that courts have failed catastrophically in some cases and improvement in how they're conducted or how prosecutiorial incentives are placed or examining other aspects of the system would be worthwhile, I heartily agree.

But if your point is that courts are useless when it comes to arbitration of the truth, then you're not on any particular road to truth. Social arbitration of truth is necessary for society. There have to be some institutions trying to do that work, whatever shortcomings they have.

Courts are one particular model for such institutions. There may be others but most institutions are going to converge on similar functions (though potentially distributed across different roles with different incentives).

They do rule on factual matters. But it has nothing to do with the truth (well, we hope it has something to do, but there's no guarantee whatsoever), it's just legal action. They can rule sun rises on the west, but it won't change the movement of the sun. They can rule somebody is a murderer, but whether it's the truth or not is not determined by that ruling. It could be either. The legal consequences are determined by the court ruling, but the actual murder doesn't change. We hope that in most cases the ruling coincides or at least resembles the truth, but we know for a fact (and this is the truth) that is it not always so.

> Every human being and every institution has such limits

That's exactly the point. So we shouldn't act like these limited institutions suddenly gained magic capacities because we conducted some complicated procedures involving some highly paid people.

> But if your point is that courts are useless when it comes to arbitration of the truth

No, they are not useless. They just aren't the truth, they are implementation of a flawed process that hopefully leads us closer to the truth than any of the other the alternatives we could find. It's not binary - it's either whole truth or useless.

> There may be others but most institutions are going to converge on similar functions

There are other models. I wouldn't claim they are better or worse, but they exist. I would recommend checking out: https://www.goodreads.com/book/show/30066446-legal-systems-v... Are any of those better than what we have? Who knows. Maybe we had the luck of finding the best one. Maybe in 200 years history enthusiasts would be laughing at us idiots subjecting themselves to such obviously flawed and useless system.

And yet no indictments for anything resembling collusion, i.e. coordination between the Trump campaign and foreign actors.

The fact is that liberals and the media subjected us all to 2+ years of peepants reporting, telling us that all of the shoes would be dropping any day now, chasing a fever dream of a grand conspiracy with Russia. And it was all total horseshit, despite the fact that the NYT, Wapo, MSNBC, CNN et al filled volumes with reporting on the subject. These are the loudest voices in the country, they're supposed to use facts to buoy the public and keep us moored to reality, and instead showed themselves to be the most detached.

If you can't even acknowledge this basic state of affairs, then we're not even talking about the same universes and there's no conversation to be had about the factual accuracy of the media.

This is inaccurate. Go read the report, or at the very least read the Wikipedia summary.

The conclusion was that a conspiracy between Trumpworld and the Russian government could not be proved because the potential suspects on the US side hid, or destroyed, evidence and lied to investigators.

What were these people covering up -- at considerable risk to themselves for prosecution -- if there was nothing to hide?

Disinformation is dangerous because too many overly credulous people are all to willing to believe in alternative facts even if contradictory primary source materials are available.

The talk page for that wikipedia summary is also entertaining.

If and whether anyone impeded the investigation is beyond the point -- there is nothing in the Mueller report about,

- money laundering through Deutsche Bank

- quid pro quo for a Moscow tower development

- secret mail servers

- Trump as a "Russian agent"

- incriminating piss tapes

And many other topics which don't immediately come to mind but can be easily googled. For two years, these are the stories we were all told were just about to drop.

If for those years, the claim had been "Trump and Russians may have worked together to take out Facebook ads, push bogus articles through troll accounts on Twitter, and phish Clinton aides" nobody would have given a shit. Yet even this set of underwhelming claims probably exceeds all that the Mueller investigation suggests could allege to have taken place at maximum.

You're shifting the goal posts from the question of a campaign conspiracy (coordination) to a broader question of inappropriate connections between Trump and Russia.

Further, the purpose of the Mueller report was not to investigate any of these issues. The Mueller investigation was very narrowly focused on whether the Trump campaign conspired (not 'colluded') with Russian entities during the 2017 campaign.

The additional issues you raise were outside of Mueller's remit. Those questions were supposed to be answered by a separate counterintelligence probe conducted by the DoJ. This separate investigation has vanished without a trace[0] and appears to have been killed or suppressed.

Leaving aside the conflict of interest issues inherent in the Trump DoJ investigating Trump, why would this investigation disappear, given how eager Barr would be to exonerate Trump if the outcome of the investigation would give the DoJ the slightest figleaf of justification to do so?

[0] https://www.washingtonpost.com/politics/2019/05/15/what-happ...

This the finest special pleading, and exemplifies exactly what I'm talking about. It's fine for you to insist on this very narrow and technical view of the situation, but that's absolutely not how the case was presented in the media day and night for years, and I have no desire to engage you on these minute points.

I'm not going to argue with you about these specific facts, because the correctness of facts composing your take does not even matter for what I'm saying. You want to argue about the specific gastronomic precedents of a shit sandwich, whereas I'm telling you that the problem is that everyone was told for years that a gourmet hamburger was on its way. It doesn't seem like you have anything to say about this state of things, which is why you keep dragging the conversation down into how many kernels of corn did the sandwich contain, and of which variety.

The Russian conspiracy story did not materialize anything remotely living up to the fantastic sizzle -- a fact made all the more amazing by the staggering abandonment of journalistic principles undertaken en masse by establishment media in promulgating that story. Nobody can be taken seriously who thinks these same people are arbiters of fact.

I'm genuinely impressed with your language and use of culinary analogies. I wish I could learn to write this way

Facts aren't lies, and facts are definitely not opinions.

Hate speech absolutely exists. Rights end when they infringe on the rights of others.

You can easily do hate speech without telling any lies, just presenting [a specific subset of] verifiably true facts.

Hate speech definitely exists. A lot of people, regrettably, hate other people for a variety of reasons, and speak about it. But there's no "right" to be loved, so neither speaking nor hating violates any rights.

And this is why we are tormented with free-speech absolutists - their conscience says they're doing a good thing for the world.


Defining things as "other's freedoms" is begging the question. We generally recognize, for instance, that no one has the "freedom" to drive cars as fast as they like or fire bullets in public. Not many people think they have the "freedom" to knowingly tell lies about me or to reprint my copyrighted works.

My existence is quite sorry, sure, but it would be improved if my ideological opponents could make arguments with substance.

> fire bullets in public

You can fire bullets, provided you can guarantee you wouldn't hit anyone or destroy any property. Usually this is achieved by either doing it on a range or in an open space far from other people and their property. Same for driving - if you're driving in a desert where you can't hit anyone, you can drive as fast as you'd like. If you're on a road surrounded by other people who you could maim or kill, and/or destroy their property - that's another business. It's not because I feel "offended" by you driving too fast - it's because you literally could kill me if you drive too fast. I haven't heard of anybody killed by a facebook post that expressed opinion they didn't agree with. That's why this is not your right to not see such opinions.

> but it would be improved if my ideological opponents could make arguments with substance.

Maybe it would, but nobody owes you to improve your existence. The system where certain people were forced to improve other people's existence, whether they liked it or not, existed once in the US. It was called "slavery". After a while, people wised up and realized this system sucks. At least most of the people did. Since then, you owe everybody not to make their existence worse by violating their rights, but they do not owe you to make your existence better - not until you convince them to do it by offering them something in exchange, like money or hugs or whatever you agree upon. It sucks, I know, but better than the alternative.

> The system where certain people were forced to improve other people's existence, whether they liked it or not, existed once in the US. It was called "slavery".

I didn't realize that the US never had taxation, the draft, nor jury duty.

This problem is far deeper and more complex than the author makes it out to be. I have yet to see a single detailed explanation that actually touches on all the intricacies involved: from blog posts to documentaries, they all are incomplete at best or wrong at worst.

So of course, the proposed solutions are necessarily incomplete, as they're not working from a comprehensive understanding of the problem. I don't know how to solve this. The problem isn't recent, it's at least decades old, if not centuries. As far as I can tell, it boils down to a lack of effort to reconcile the experiences of disparate groups, and the need to contextualize their ideologies within those experiences. This fundamentally means that one person's facts can be another person's fiction, and the line between the two should be tread carefully. For a long time, this problem was sidestepped entirely, as groups simply avoided one another or had minimal contact.The problem was exacerbated in the modern world, and has been brought to its most extreme conclusion through the internet.

Institutional authority today speaks with too much unearned certainty. Those who share that certainty will be appeased, and those who do not will be incensed.

I do not know how to solve this problem. Neither does anyone else, though. That includes the author of this article, who doesn't even understand the problem in the first place.

If there is one takeaway, however, that I would like people to have, it is this: just because someone disagrees with you, or refuses to acknowledge facts of which you are certain, does not mean they are stupid. It is entirely possible it is you who are missing something.

Be uncertain.

The solution is simple, yet unoptimized entirely but for the smallest union of people: Empathy

The incentive everywhere need to be to empathize and understand our differences, to bridge the gaps. When one looks at the overall business world, which is institutionalized war and injustice, against people and environment, there is much to work with.

All the logic in the world won't right these wrongs.

Most people don’t have that much empathy. Most of them don’t even want it. (I certainly don’t want it.) Your solution is basically adding terms for other people to everyone’s utility function. It would require a super-Orwellian effort to pull that off. The truth is that there is no panacea at all, and we need to design sound game theoretic mechanisms for every single thing we care about.

With empathy comes many gifts, well-being, and a culture that is more sustainable.

"Fact-checking" has very little to do with facts. It's just a clever rhetorical ploy to promote some viewpoints and disqualify others. Specifically, it's a species of bullshit[1]. If the actual facts support the bullshitter all the better, but if they don't you can rest assured the "fact check" will be "contextualized" to whatever extent is needed to serve the bullshitters' rhetorical ends.

[1] Bullshit defined here as speech intended to persuade without regard for veracity. The liar knows the truth and conceals it, but the bullshitter doesn't care if his statements are true so long as his bullshit is persuasive. https://www.goodreads.com/book/show/385.On_Bullshit

This is how I feel about people claiming every dissenting view as "fact checked" or "debunked" in recent times; it just throws out any possibility of debate.

Basically, you've decided the fact checkers are somehow malicious actors, without any particular evidence of that actually happening, and vaguely labeled their speech as bullshit.

Seems like nonsense[1] to me.

[1] I mean this as the word is normally used.


> Mixture: During the debate, Trump was repeatedly asked and invited to explicitly and clearly condemn white supremacists, and he did not do that.

He did condemn it; it literally follows up with "Sure, I'm willing to do that."

And for added bs


38 Times President Trump Has Condemned Racism and White Supremacy



> Mostly False: Biden said "Now we have over 120 million dead from COVID" at the Pennsylvania event in apparent reference to U.S. fatalities.

There is literally a video of him in the article doing it


Took literally a couple of seconds to find evidence.

You're trying to fact check me. If we apply the logic of the post I replied to, that means your post is a rhetorical ploy and bullshit (see prior footnote 1).

Oddly enough, I'm not sure there's any sensible way to refute me without first trying to refute that original post.

> Basically, you've decided the fact checkers are somehow malicious actors, without any particular evidence of that actually happening, and vaguely labeled their speech as bullshit. Seems like nonsense[1] to me.

^ This is you stating that fact checking isn't used by malicious actors and there is no evidence to show it is.

I just demonstrated with evidence that it is.

You are now trying to twist your way out of this by equating "fact checkers" (IE major organisation presenting opinions as facts that are not up for debate) and the act of correcting misinformation as one and the same thing.

Twisting of words is one of many dishonest tactics used by progressives in debate and it's the same one you are attempting to use right now.

For the record, I'm good at pointing out dishonest tactics and people who use them in debate, I wouldn't continue.

Also while this is going down this path, I'm going to bring up the narcissists prayer so we can refer back to it later if you continue with this approach.

You could indeed see yourself as being "fact-checked". But this doesn't mean the person attempting to persuade you of their point of view thinks of themselves as a "fact-checker", or popularises themselves as a source of information that is always more correct than any other.

I appreciate that you did, in fact, try to argue against what I said without the use of facts, resolving the paradox I presented. If with a somewhat strange arguement.

I appreciate your good-natured discussion, but disagree: my statements were factual, and it's not strange to present hubris as an indicator of untrustworthiness.

I read it as: self-annointed arbiters of truth demean themselves by their own definition as such.

Trustworthy sources of information accept the limits of their knowledge and recognise their own ability to make mistakes. Who fact-checks the fact-checkers?

"Fact-checking" by definition, implies infinitely-superior, infallible knowledge.

Ie, bullshit.

So you believe nothing is a fact?

You're obviously not the target audience of fact-checkers. Most people agree on some basic principles of reality that can't be skewed.

For example, when a politician says something in public that millions of people hear and is recorded on video, and then he says he didn't say it, that is a lie. A fact-checker would provide the evidence and context.

Or if a politician says that nothing could be done to reduce Covid deaths in the US, that is less straightforward but also an easy lie to disprove because there are people who died due to lack of available tests for their coworkers, or people who died because of a huge nightclub being allowed to stay open.

> you believe nothing is a fact

No, but I do believe that “fact-checkers” get it wrong often enough to be skeptical about what they say is.

Any citations for that one? Everytime I see a fact checker called biased for one way or the other, not only is it usually salty people of the opposing views usually saying it, it always seems like tittle-tattle rather than actual systemic bias.

One that recently really bothered me was a black box on YouTube videos that mentioned anything about the US election. The black box said that "the AP has called the election" -- the Associated Press is not responsible for determining who governs the US.

Also, for what it's worth, it was a math video I was watching.

> The black box said that "the AP has called the election" -- the Associated Press is not responsible for determining who governs the US.

You don't seem to understand that this is how US elections are always called? The press calls the state when they estimate it's impossible for it to go any other way, and then the electors vote a while later.

It's the presentation of a single non-governmental, potentially biased organization, the AP, as the authoritative figure, to which I object. The black box message didn't mention electors or counts at all, only that "the AP has called" it.

Whether or not the election is called is the only thing that mattered the week of the election since there was/is no finalization until the electoral college votes. I don’t understand why there’s any handwaving about this part other than the sudden encouragement by certain political actors that never had a problem with it before.

Oh, that's fine, it's that a specific commercial organization was presented as the deciding authority that is the problem. That won't help convince anyone whose fundamental disagreement is a distrust in traditional media's authority figures. Google/YT could have put anything in that message, and they got the delivery all wrong.

If the problem is fundamental distrust then what difference does it make if it’s the AP? The message isn’t for the people who choose to be misinformed, it’s for those that could be unwittingly.

I hope you agree that context matters. Anyone can have a sound bite cut out of a larger statement and have that spun to be something that isn't representative, or that he didn't intend, or even the complete opposite. And then for many people, repitition becomes fact. Whatever they've heard most often, from the most sources, becomes their belief.

For something to be fact-checked, it must take context in consideration. Fact-check can be true, a middle ground (both true and false) or false. So there's not this black and white as some people pretend it is. A few just want to pretend they should get away with lying, and getting people to go along with their villainry. Beware such sociopaths, they will do the same to you and loved ones at next turn. Never yield any power to such people. No cause is worth destruction.

If your statements don't hold water, you should be held accountable for it, and there should be consequences according to your power. When those consequences is more power, you sooner or later arrive at Mussolini (look him up).

Though, for many people, facts, science, teachers, doctors, none of that matter. So this means the political system must get out of its stalemates and address the real issues that lie beneath the bullshitters. People don't think alike, so you can't reach everyone, even if you think you have all the facts. Besides, facts can always be disputed, though not unreasonably so.

My conclusion is, no fact-checking will ever be sexy and viral. People will die of boredom instead :)

Good try though!

I recently noticed truthorfiction.com being prominently displayed on Twitter/Google searches related to the election, so I looked into who's actually managing content there. A few quick Twitter profile searches on the "Our Team" people revealed years of scathing Tweets about Donald Trump.

Do you have an example of what you're talking about?

A number of fact checks just amount to "don't say it like that, say it in a way that supports our narrative instead" when there's no factual error. E.G:

https://i.imgur.com/EmalFyS.png https://i.imgur.com/bBCOXzv.png https://i.imgur.com/UZam8qQ.png

In cases of no evidence, they'll usually rule favorably if it's a narrative they like. Such as here with "Unproven, but alludes to a deeper truth":


Compared that to consistently rating Trump's claims as "Pants on fire" for lack of evidence, or "Mostly False" even when they found strong supporting evidence:


There's many topics which, even if I agree with the writer, should be in the realm of an article or opinion piece rather than a fact check (E.G: "AP FACT CHECK: Trump distorts record; BLM falsely accused of coordinating violent protests"). Fact checks should ideally stick to debunking falsifiable claims, rather than providing a counterargument for/against a large political issue.

I think the biggest problem is that much is said to be fact-check but is, indeed, opinion verification.

People have many degrees in how they express themselves and many times an opinion use words that resemble factual statements. Then others read it as facts and want that someone give a validation or rebuttal of it.

Add some partial fact-check, some political inclined, some strong words, and then everybody has opinion of the credibility of it. Be it centralized or distributed.

I think the key to achieving this is building Web of Trust on top of the fact checker relationships and connect them to the users, so that you can see exactly how you are connected, and whether you can trust the individuals.

Today, the sources of whatever information you read, even from the most credible sources, are obscured. Whether it is CNN or a peer-reviewed journal article, the most information you will get about the author is a name, a picture, and maybe a bit of their publicly available data.

With Web of Trust, the level of trust you should assign to an author, and by extension the information, can be instantaneously computed by adding up your own trust references.

If I assign my trust responsibly, and you assign your trust to me, then everyone I trust is by extension trustworthy, and the information metadata can reflect this.

If you don't think I assign my trust responsibly, then you can remove me from your trust circle, and my assigned trust will become irrelevant.

It's not a perfect system, it's vulnerable to attacks, it requires eternal vigilance, but it's the best we've got going for us today, IMO.

I'll drop https://spritelyproject.org/ here.

Having said that, there is likely to be an issue of influencers ie. Those celebrities that you follow as they're interesting / exciting etc but may be completely uninformed on say green issues, which then have an undue influence on their followers. Worse, many of them are sponsored and corruptible that way - we're talking about boiling frogs both for the influencer and followers basically.

The other issue is that a Web of trust has a tendency to echo chamber. I think the actual hardest thing to do is to cross-polinate networks in a voluntary way, that I don't know how to do yet.

This system reminds me of a sophisticated Facebook news feed. Though a major problem is the news sources that are biased at the source. We might need a transparency fact checker that compares between sources.

Here is a design proposal for such a system with an interesting twist: https://www.lesswrong.com/posts/ZdtFBCtixqay5aoWF/design-tho...

I'm working on building this. Email me if interested! toomim@gmail.com

I sent you an email, did you receive it?

have you used a web of trust system you liked, that works well w/ an existing social platform? I'd try one if you have a rec.

fwiw I wrote one of these and mine sucked

I'm curious-- did you build it as a decentralized/federated/distributed piece of software?

nope -- it was a forum, the web-of-trust feature only operated over the users and content inside the forum

Could you please write a little more about your forum?

I'm also working on a forum system that includes basic WoT features, and would be grateful for anything you share.

(Prototype link is in my profile if you want to play with the current prototype.)

I second that!

I'm glad to hear OP at least attempted WoT inside a centralized forum. I think one of the problems with WoT is that many projects hard-couple it to a distributed approach to software that utilizes it. Since OP didn't take that approach I'd be curious to know what didn't work.

it wasn't a design decision to link the WoT to a single forum -- I was just playing around and didn't want to go through the hoops of integrated with an external social network

Mine is here -- https://github.com/abe-winter/trustgator

It's still very much a pre-MVP prototype, but try the link in my profile.

I'll go further. We need at least two webs of trust.

Would you trust a Democrat to fact-check Trump's claims of election fraud? Would you trust a Republican? In fact, you should trust the Democrat if they say the claims are true, and the Republican if they say the claims are false.

Ideally, we'd get one set of fact-checkers that both sides trust. In practice, that may not be possible. But at least, if we have separate fact-checkers for each side, if they agree, then we have a fairly good basis for suspecting that they're right.

What do you do when the Republican fact checker repeats what the leader of his party tells him to say even though it is fact-free?

A majority of Republicans believe Trump won the election (!) because that's what Trump is saying. There is little reason to think a Republican fact checker would contradict Trump. Indeed, any fact checker who did would be ostracized as a RINO, much as has happened with Fox News since they acknowledged the outcome of the vote.

Partisan fact checkers would only give the Republican party a safe space to continue to lie about matters of objective fact, such as who won the last election.

Frankly, the fact that 'what is truth' can be framed as a partisan issue in America speaks to much, much deeper societal problems than organized fact checking can fix. This is an attitude that is much more in keeping with that found in countries like Albania or North Korea rather than in another advanced democracy.

Please see my comment where I addressed this:


> There is little reason to think a Republican fact checker would contradict Trump. Indeed, any fact checker who did would be ostracized as a RINO, much as has happened with Fox News since they acknowledged the outcome of the vote.

Those two sentences contradict each other. And the second makes my point. When Fox News says that Trump lost, you can probably trust that particular fact-check. ("Confirming testimony from a hostile witness", in law-speak.)

Since Fox News started fact-checking Trump, Trump's followers have stopped treating Fox News as a reliable source and are voting with their feet towards OANN another very far right sources. They're clinging to lies (politely: disinformation) even in the face of evidence to the contrary from a source that they formerly considered reliable.

That is my point.

Conservatives (or, at least, Trump followers) will turn their backs on sources of information that they considered trustworthy if those sources contradict the leader de jour.

For most people outside US, like over 7 billion people, both of these would be invalid webs of trust.

I wouldn't trust anyone who uses partisan lines to assign trust. And, in a transparent WoT system, I would be able to see them doing this with an easy SQL query. I would then remove any #trust in this individual, and this particular WoT's trust level would be worthless (in my graph.)

That's a good point. I think more than one Web is a great idea and desirable.

As far as your questions:

I would trust some people who identify as Republican, but not others. I would trust some people who identify as Democrat, but not others. Out of the two sets, I would trust the ones whom I personally know, and not anyone else, no matter who they say they are.

In the following scenario, I would trust both a Democrat and a Republican fact-checking Trump's claims of election fraud;

[Me] #trust [My friend Alice, whom I trust to assign trust responsibly] #trust [Alices's friend Bob, whom Alice trusts] #trust [...] #trust [Democrat fact-checker] #approve fact-checking report.

Depending on the Bacon number and how many other similar connections there are, I would trust them more or less.

If there are no connections, then I wouldn't trust them beyond the distance I can propel their supposedly human body under my own power in Earth's gravity.

Recycling my comment from a recent discussion: The problem is that we don't want facts - we want confirmations.

"The reality is that the implication of the Internet is that ideas are in abundance, and people will seek out what they already agree with, as opposed to accepting what is delivered to them." https://stratechery.com/2020/the-idea-adoption-curve/

Fact checking is not the right angle to look at it - what we need is curiosity.

"fact-checking" is dog whistle for "only leftist allowed"

The role of fact checking is not to strive for partisan balance but rather to identify public comments that are untrue.

American conservatives would fall afoul of fact checking far less frequently if the conservative worldview wasn't wedded to talking points that are not factually accurate. For example:

There is no evidence of widespread voter fraud in the 2020 presidential election. There is no evidence of ongoing large-scale voter fraud in US elections. There is no evidence that the incoming Biden administration intends to end capitalism. There were no 'convoys' of migrants preparing to violently storm the US-Mexico border prior to the 2018 midterms. Barrack Obama was not ineligible to serve as President. There is no evidence that the Democratic party is detaining children in the basement of a pizza parlor. There is no evidence that the Democratic party is run by a cabal of pedophile cannibals. Etc.

A fact checker who does not point out the fact-free basis of these, and other, deeply-held conservative talking points is simply not doing his/her job.

If the weight of fact checking falls primarily against conservatives, then conservatives can fix that problem by lying less frequently.

What about lefty antivaxers?

I'd like to know how distributed fact checking could deal with this example: trump's infamous comments about mexican border crossers including rapists and criminals.

From the quote, it is clear that the context is people crossing the border are frequently criminal elements, rapists being some of them. But every time I hear reference to this quote(except from overt trump supporters, who have their own issues interpreting events ), it is characterized as trump saying that all Mexicans are rapists. Literally every time I've ever heard it referenced, even from mainstream news articles.

How does one do distributed fact checking if there is a mass public delusion? At best, distributed fact checking can give you th concensus of fact checkers, which is not always the truth.

what percentage of immigrants - of any origin - are criminals and rapists?

the quote:

> “When Mexico sends its people, they’re not sending their best,” he said in the same speech. “They’re not sending you. They’re not sending you. They’re sending people that have lots of problems, and they’re bringing those problems with us. They’re bringing drugs. They’re bringing crime. They’re rapists. And some, I assume, are good people.”

you can assume he meant something other than what he said, but his words were in reference to "When Mexico send its people", so it is not "delusion" to say he referred to "people from Mexico".

he so generously acknowledges that only "some" of them might be good people.

what do you think he meant when he said "they're not sending you"?

what do you think he meant by saying Mexico "sends" people to the US?

the media was right to treat that comment as indefensible.

It being indefensible is a different point than characterizing the statement as "trump called mexicans rapists".

You're doing the exact thing I was saying that everyone does with regards to this comment.

Children need to be taught about fake news, bias, etc.

May not necessarily have much of an effect, as in this classic example where children who had recently received education on the use of fake sources were unable to spot that an article on the "tree octopus" was made up:


There is very little evidence that fact checking actually works to reduce the impacts of disinformation. See e.g. this study of the impact of Twitter warnings on belief in disonformation:


The only known-effective way to deal with disinformation is to remove it.

Democratized fact-checking won't work. The costs are too high. IF you can find a way to reduce the costs of the general public to fact-check, you might be onto something. But until then, we'll see a few places that sometimes provide signal to the general public but mostly provide noise. This problem isn't simple. If you have ideas, pitch them and build.

Ceding "fact-checking" (i.e. narrative construction) to social media companies (directly, or indirectly by letting them choose contractors to do it for them) is a terrible idea. Not only is it unlikely the narrative you receive will track reality in a useful way, but it gives a huge amount of political power to these companies.

A much more sane approach, IMO, is to come up with software frameworks where you choose which research firms or individuals (friends, professors, etc.) you want to follow and how much trust you want to assign to each of them. There is plenty of research on ways to do this well in the context of "web-of-trust" cryptosystems.

I wonder if pulling the brakes on what's being called the post-fact era, is futile. Maybe the internet is perfectly capable of maintaining N disjointed realities and there's nothing we can do to resolve them.

It may not be high-resolution virtual reality, but if the only way I interact with other people is through devices, it's virtual reality. We call them bubbles but they're so much more.

I think if we accept this as an invariant we might be able to find solutions by trudging forward. As a thought prompt it certainly gets my thinking cap going.

The problem is that these coexisting fantasy worlds run into issues when you actually do touch reality, which does not go away simply because one choses not to believe in it.

covid of course being the most recent example. We can all run into whatever bubbles we want, but it won't stop a biological virus. Everyone can spin it into their own virtual narrative, but people will die nonetheless, and underlying the cloud and the internet is good old hardware and infrastructure, which is very real and very decaying. Fleeing into digital worlds is only a temporary solution.

That’s a very compelling thesis statement. I’m looking forward to the discussion here.

What criticisms of your argument are you anticipating, if any?

"Obviously nobody should be allowed to write online and moderation just postpones that desirable end state"

would be a valid counterargument I think

I'm completely convinced by now that fact-checking is utterly useless, because it is based on a specific worldview that truth is something important.

While I happen to share that worldview, recent populist politicians didn't just ignore that, but managed to turn this idea on its head — and the left still seems to not get it.

There was really no shortage of scandals, proven disinformation or exposing books. But I saw liberals and progressives stumped over the fact that exposing those lies did absolutely zero to change opinion over the populist administrations. No matter the previous results, all those years they still worked with the mindset that just exposing one more lie would suddenly change everyone's mind.

As Einstein said: The definition of insanity is doing the same thing over and over and expecting different results.

The thing they don't get is that the average Bolsonaro or Trump supporter simply does not _care_ whether their politician lies, if they simply come out of the argument as stronger. It's not logical at all, it's psycho-logical.

Any facts are pretty simple to get off the table if you're brazen, narcisstic and sociopathic enough: Just accuse the other side of doing it before they do, and if they do, double down on it yourself in order to keep the narrative lead.

It doesn't even have to make the slightest sense, it just needs to make small egos feel good and in the know.

"ABC: I'm lying? Look at XYZ, he lied two times and I can prove it! So nasty!"

Suddenly the distinction between ABC lying every day and XYZ a dozen times in their career does not matter anymore to the casual onlooker. All politicians lie, why would fact-checking matter? In fact, probably even the fact-checkers lie!

And if you don't believe me, just remember that "Fake News" was an invention of the left that was rhetorically completely spun around (extremely successfully so) from meaning creating fake news to meaning that all (mainstream) news are fake.

It took nothing more than repetition and massive confidence. As long as you keep people more scared about China, socialism, gun-robbers and making the next paycheck, it's rather easy to control the narrative, no matter the facts.

Fact-checking is a liberal pipe-dream.

Fact checking works in a civilized society where the adversaries are “good sports” who acknowledge the rules of the game — e.g., conceding a lost election. The problem today is that we don’t live in such a society. Our political reality has edged closer to pure wrestling for power, where the other guy is seen as bad and any lever is ok to pull to prevent them from expressing.

Who is funding those goons?

Very long article that buries the lede:

> We need a market ecosystem of moderation vendors who provide different kinds of value to different kinds of users. A moderation marketplace is good for platforms because they don’t want to be the arbiter of truth, they just want users to have a good experience

> It’s good for users because they’re not forced into anything; they can pick which web they want to dwell on. This may sound like a filter bubble, where ‘everyone has their own truth’, but moderation vendors will still need to maintain their reputation for being right

The author is overly optimistic in thinking that people who fall for blatant misinformation, will not fall for grossly biased or incompetent moderators. Anyone who believes Trump's election-was-stolen conspiracy theory, will also have blind faith in any moderator that Trump and his movement endorses, regardless of the moderator's reputation among mainstream sources.

You may say that doesn't matter because the vast majority of people will never tweak their settings and will simply use the platform's default moderator. Which is a good point, but once again, that brings us back to the status quo where people get upset about the "bias" shown by the platform and its anointed default moderator. I do like the idea of decoupling the platform corporation and the moderator, but it still raises questions about who gets to control the moderator organization.

Call me crazy but these misinformation campaigns have made one thing abundantly clear. Popular voting is inherently susceptible to actively-misinformed voters and there are no palatable solutions to this problem. https://outlookzen.com/2020/10/26/reasons-to-replace-the-pop...

The real basic failure is the same one which is currently running: how do you limit it to 1 "author" per actual physical human being in the world?

Like set aside other problems like establishing credentials or reputability, disinformation campaigns are based on the reality that it's very easy to create just plausible enough looking accounts, and getting easier (https://thispersondoesnotexist.com/).

The only ones who need agonize over the moral quandaries are the ones setting censorship policy. For the rest of us, it's helpful to stop worrying about ought and focus on what is. In the long run, an audience will leave a platform that discriminates against them. This means the big platforms will fragment. The more they discriminate, the faster this will occur. This is a financial and social benefit for their opponents, who should therefore expose but not oppose it.

Monolithic institutions such as Big Platform are vulnerable to the same ideological convergence that captured the big universities, but this problem is self-solving. Social cycle theory predicts hegemonic collapse due to impractical religion, but now both universities and Big Platform are being replaced piecemeal by the invasive barbarian hordes of the decentralized Internet, who ignore national boundaries.

See Parler and Gab for Twitter; DDG, Bing and Dogpile for Google. Facebook's challengers are VK, Minds and Diaspora, although Facebook is more loathe to offend its conservative lucrative userbase. Google's manipulations are subtle enough that anti-trust, section 230 and the like are probably greater threats; personalized search results still usually deliver a superior product despite biased results manipulation.

A divided society that increasingly refuses to do business with each other cannot have united platforms. For every leftist who discriminates against the right, there is a rightist who is legally barred from openly practicing his discrimination against a protected category on the left. Which isn't the same as saying this discrimination doesn't occur.

Already hordes of Republican Boomers have rejected Twitter's heavy-handed censorship, driving record competitor growth. Regardless of whether one believes Trump will overturn the MSM-declared electoral outcome, it's obvious that IF he does, Twitter's losses will accelerate. This will increase popular skepticism towards the other big platforms, to say nothing of increased governmental scrutiny.

Hillary's loss proved certainty an illusion. It's a bold bet; we'll see what happens.

Just wait until they start "fact checking" your personal conversations. Yet another unfortunate combination of elitist and philosophical dunce. It is so tiresome.

You assume the average person cares about facts more than validation of their preexisting worldview. Fake news peddlers know this isn't true, and it's why they're winning.

I think "fact checking" is the wrong framing for the problem at hand - assuming we're talking about the problem of people shooting up pizza shops. (The article also talks about how businesses or hobbyists who are genuinely interested in truth can find the truth more accurately, but I think that's pretty orthogonal to the problem of social-media-fueled conspiracy theories.)

Take the US election, for instance. At this point the "facts," strictly speaking, aren't really in doubt: we have certain reports from the various states following various procedures for verification, and those reports give a certain winner. The argument against that position is not against the checkable fact of the reports - the argument is that, for various reasons, multiple of these states are either deliberately lying or had their procedures tricked. What's left is whether you believe that to be plausible and in fact more plausible than the null hypothesis that the election went roughly as predicted. And that's not checkable, or at least not democratically-checkable. If you believe that voting machines were hacked in a way that left no evidence, or ballots were modified in transit, or the distance of observers left them unable to notice intentional mis-tallying, or whatever, what sort of fact would convince you otherwise?

If you are so inclined to believe that the election was stolen, you can find facts to back you up. It is factually true that there were illegal votes, for instance: in an election with 160 million votes, the chance that not a single person would misunderstand the rules for where they can vote are basically nil. It is factually true that the Georgia secretary of state made this argument, not particularly eloquently, and is on video saying that they expect to find voter fraud. It is factually true that many electronic voting machines lack proper safeguards. It is factually true that there are more avenues of attack on mail-in ballots. And so forth.

What we have is not a lying problem: what we have is a propaganda problem.

And we do. in a sense, have "distributed fact-checking" / a marketplace of fact-checkers: we have different members of the news media with different viewpoints. And it turns out people aren't particularly loyal to any one vendor. When Fox News called the election for Biden, people flipped the channel to Newsmax, which has "an editorial policy of being supportive of the President," according to their CEO. So I don't think that encouraging people to buy into fact-checking builds any bridges for the communities that incubate future pizza-shop shooters to come back to reality. What they will learn is that they can always find facts to fit their beliefs. Sure, there are facts out there that don't fit their beliefs, too, but they lack an interest in changing their beliefs, so it's easy to wave those away. (The flat-earthers mentioned in this article are another good example of this phenomenon.)

If Donald Trump had just spent 30 minutes more a day reading books or briefings, we wouldn’t be litigating misinformation, disinformation, flat earthers, fact checking, conspiracies, boomers, etc.

Thousands of rants never written, tens of thousands of podcast episodes would have never been recorded, millions of tweets never tweeted.

You actually only needed to improve the literacy of one man to make irrelevant 99.99% of the problems described here. Under the Biden presidency all this shit will still exist, it won’t disappear on noon Jan 20 2021, but it will stop mattering. Treating Trump as a constant while blaming everyone else (like this does) is the trap.

This (extremely dubiously) assumes a sincere effort to tell the truth on Trump’s part

> Under the Biden presidency all this shit will still exist ... but it will stop mattering.

Why is that? Because Biden is truthful? Because he won't be tweeting 10 times a day? Because the media will (effectively) do the tweeting for him? Because the media will simply not challenge him, and ignore his critics?

All politicians lie, or perhaps to put it more charitably, they promote a very one-sided point of view. It's almost the definition of politics. The only reason anything will change under Biden is that many of the people with the biggest megaphones support him.

News has been trump trump trump for the past 4 years and you say he doesnt get his voice heard?

The nature of the modern right is a perpetual persecution complex. It doesn't matter that we just finished 4 years of a republican president with two of those years including a senate and congressional majority. According to their rhetoric they're still the voiceless, powerless underdogs persecuted by "the system".

"Vast right-wing conspiracy"

It's politics, as I said. Two sides of the same coin. It's not a right/left thing.

Exactly. People get lost in ideology, and lose their principles. Left is just a reaction to Right, and vica versa. Both parties hostages in a two party system of stalemates.

These political axis are defunct in the rest of the world too. They serve only the purpose of career politicians. And no, the "rebels" aren't the "final solution"..

There's no question Democrats have a stranglehold on our education system, news media, Hollywood, and all major tech companies. That's what drives the feeling of frustration -- of being voiceless. (The left has rightly told America that representation matters, but it only seems to matter on identity axes. Meanwhile 50% of America has no newspapers, few schools, no social networks, few entertainment options...)

Besides, there is no shortage of left-wing persecution complex. It underpins so much of the woke movement it's been given it's own neologism: grievance olympics.

If it's a call for platforms for lying, it is not welcome.

I don't see how you come to that conclusion. I didn't in any way say that Trump has not had his voice heard.

You assume that Trump is not just deliberately lying. While Trump might or might not be smart, he is surrounded by people like Pompeo and Cotton who certainly are.

They lie and mislead too.

I don't think it'll go away, it's still going to be a huge problem. But Trump was definitely a super-spreader.

> If Donald Trump had just spent 30 minutes more a day reading books or briefings, we wouldn’t be litigating misinformation, disinformation, flat earthers, fact checking, conspiracies, boomers, etc.

He was elected to lead, not to read!

He's following the sociopath rulebook. He knows what he's saying and doing. First step is the charm stage. Hell, he's the top boss and complained about election fraud months before election even started. He did nothing about it other than sabotaging US Post. Later he latches unto his anchor to defraud half the population, to spread FUD and lies.

The fact people can state this without reprecussion, makes him a weak liar. Everyone knows this, and this is how he speaks first, in order to disarm allegations about what he's doing. Just project all his words back to him, and you now have a list of all wrongdoings, ironically.

Unfortunately, in the US, fact-checking addresses only the symptom, not the cause, and is unlikely to change anyone's minds [1].

We can see that disinformation isn't a necessary feature of democratic debate today, as it is a tactic used overwhelming by Republicans/conservatives, and very little by Democrats/liberals.

In a two-party system like the US, the structural reasons Republicans adopt this tactic is fundamentally because the structure of the Electoral College makes it rational for Republicans to target polarized extremists (to boost turnout), while Democrats need to target the center (to win swing voters) [2].

The solution isn't fact-checking. The solution is abolishing the electoral college, so that both parties have to appeal equally to centrist swing voters, who are more convinced by reasonable factual appeals than by polarized disinformation.

[1] https://en.wikipedia.org/wiki/Confirmation_bias

[2] https://www.vox.com/21561011/2020-election-joe-biden-donald-...

Nah, we're moving towards individual controls and decentralized platforms. Having a tool distribute fact check is just delaying the inevitable.

Just look at tiktok for the HARD repudiation of the ultra-left themes. The young people will always be counter-culture, and let's face it the ultra-left is mainstream today. Facebook, twitter? Those are your wine-aunts/grandpas platforms.

What are you thinking of as "ultra-left"? The only actual radicalization I've seen in the last decade has been "ultra-right," militias, normalization of anti-immigrant and anti-minority sentiment, hyper-nationalism, normalization of hate speech, hate crimes and hate speech against minorities, etc.

DHS and FBI both call out rising right wing domestic terrorism as one of the top threats to the US. They don't mention "ultra-left" anything.

'wine aunt platform' is a phrase that will stay with me

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