Fake News Challenge
fakenewschallenge.org
fakenewschallenge.org
If you don't like the way your local bakery makes their bread you go somewhere else. Now if the media reports on something you don't like someone else will tell you what you want to hear.
This issue happens on both sides of the political spectrum BUT it does seem to be a lot worse on the conservative side of things. All the "liberal" news sources like the NYT and The WashPo seem to be doing lots of reporting on Trump supporters, discussing their fears and hopes, concerns and motivations. Even more conservative sources like the WSJ, not including their editorial page, are trying to take a bigger view of the world.
On the other hand my Apple News feed seems to perpetually have a FOX News story. The one this morning was "Is Trump Bashing the new Celeb nude selfie?"
All I know is the people who seem prone to believing the stupidest stories also believe global warming is a hoax and Obama was not born in this country. A significant portion of the electorate has been drinking from the well and cannot be reached.
I would say a huge part of the problem is that news is no longer news. It's all commentary. News organizations are not only incentivized to to sensationalize stories to make money but they are also ran by people who have extreme political views and want to essentially mind control the population. They don't want to report the facts of a story. They want to make you accept their interpretation of the facts.
A great example is how different news organizations cover the recent executive order on immigration. Depending on where you get your news you were told there was a Muslim ban, a ban from 7 Muslim majority countries, only accepting Christian refugees from the Middle East, 7 countries that harbor terrorists were banned, etc.
Wherever you fall on that issue there is a news org trying to force you into a certain viewpoint one way or the other.
Please, describe to me how someone can convey pure unadulterated facts in language, without imposing an "interpretation" on them.
It's an absurd fantasy. There's no possibility for a disembodied, perfectly objective voice conveying facts. So yes, in a sense, every news org favors a certain interpretation of the facts. But these subtle variations used to be accepted as par for the course.
The root cause of the widespread distrust of media is the turn on the right towards a form of Orwellian doublespeak in their media outlets, beginning in the 90s and ramping up heavily in the Bush years. You see it's most virulent expression in the way Trump speaks: he accuses the interim AG of "betrayal" for acting on her conscious. Or the way he talks about returning government "to the people" while staffing his cabinet with billionaires. It badly subverts the ability of people on the left and the right to meaningfully speak to one another, because our words no longer signify the same thing.
Well, that's actually a pretty interesting example to look at. For one, the dossier concerned contained a lot of accusations to do with members of Trump's team travelling to Eastern Europe to meet with Russian contacts, suspect financial transactions and the like - you yourself are doing it a disservice by describing it only as rumours of a "piss orgy".
But irrespective of that, the dossier had been known about for months. But the media organisations didn't report on it, specifically because they couldn't verify the information contained within it. But then Buzzfeed decided to publish it anyway, and at that point any responsible news organisation is left in a pretty impossible position - either don't report on the thing everyone is talking about, or report unsubstantiated accusations.
Sometimes it feels like the media is damned if it does, and damned if it doesn't. Report unsubstantiated rumour? You're biased! Hold back information because it isn't verified? You're biased!
1000 * 0 is still 0.
If the Trump campaign had ordered a dossier of dirt on Hillary Clinton, and paid a respected former intelligence officer to provide it, would the claims made in it be credible?
Let's take one of the big 'fake news' items from the recent election, the one regarding Hillary's health. What do you make of this article by a former secret service agent on Hillary Clinton's health:
http://ijr.com/opinion/2016/09/260018-protected-hillary-clin...
None of the facts in that article are in disrepute either, and it's also from a respected secret service agent.
So are the questions it poses about Hillary's health more or less credible than the Trump dossier, and is it fake news or not?
The definition of fake news is getting stretched to include controversial news.
"Fake news" is fiction being sold as news on websites that look credible.
Flat out making stuff up, and polluting the media stream with it in order to manipulate people.
This is the easy case to solve. As people actually try to grapple with this, "fake news" will evolve, it will add parts of the truth to it dodge this 0th level definition.
A dossier actually compiled but unverified is still not fake news - it's a shady bad practice circa 1990. Unfortunately we're dealing with much worse problems today.
I mean that's the whole sub-prime mortgage problem all over again. Take a shitty product and wrap it in layers of respectability until people don't know any better.
When there is also evidence to suggest much of the contents of the dossier came from a 4-chan prank, it's hard to see where the difference is.
1) fake news: a recent phenomenon where fraudulent/fictional/completely made up literature is peddled as legitimate news. This is then distributed on steroids via social media. It uses unsuspecting people to transmit this.
Bias in news. This is an older problem and involves how news is presented and what spin is put. In the worst case it is deliberate in order to support an agenda.
There was a lot of anti-Trump news that basically matches what you said above:
> "fake news" will evolve, it will add parts of the truth to it dodge this 0th level definition
* Trump kicked a mother and baby out of rally
* Trump mocked a reporter's disability
* Trump proposed building database of muslims
And the list goes on.
This goes well beyond the realm of media bias, and in to direct collusion between media and the campaign.
I have no problem calling that 'fake news', which is perhaps a descriptive approach rather than a prescriptive one.
On top of that trump himself courted controversy at a rate and scale never seen before.
That's why this election was faced with drama unlike elections before them.
A lot of stuff listed above is sadly the regular calumny inflicted upon candidates. Iirc someone got kicked out of the race because of a bad photo op.
But These are all things that are problematic- without being fake news.
> * Trump mocked a reporter's disability
Do you deny that these things happened?
Has any other candidate - of any party - ever done anything like this?
Is the media supposed to pretend that it didn't happen?
This is where the "the facts have a well known liberal bias" joke comes from
For the first one, the mother who was supposedly 'kicked out' denies it: https://www.washingtonpost.com/news/fact-checker/wp/2016/08/...
For the second one, he was not mocking the reporter's disability. He was mocking the reporter for pulling a story and the impersonation he did is Trump's standard impersonation for a flustered, bumbling person. There is multiple video of him doing the exact same action to able bodied people - from Ted Cruz to military generals, and all sorts of others.
Not only that, but the reporter's disability is nothing like the action Trump did. The reporter actually has frozen/locked up joints e.g. almost the exact opposite of what Trump did.
So, either Trump knew of the reporter's disability and wasn't mocking it (completely different actions), or he didn't know of the reporter's disability (and therefore couldn't have been mocking it).
In either situation, Trump wasn't mocking the reporter's disability.
Unfortunately a freeze frame from the video of Trump's movement matches a still photo of the journalist, hence the belief that he was mocking the reporter's disability.
And this is why people say the media has a well known Trump bias.
I agree with you broadly, though. To me, fake news has absolutely no mooring in reality whatsoever and is pure clickbait - usually without political motive beyond getting angry people to click.
And that (apparently) can trace its origins to a 4-chan prank?
Personally, I do not.
http://gizmodo.com/4chan-idiots-claim-they-invented-the-trum...
You mean the ones contained in a dossier which accurately predicted several Trump related events that were not publicly known to be in planning at the time it was written, but were verifiable to have occurred as predicted before it was published?
Seems to me that it was plenty newsworthy.
BTW, it also reported that part of the carrot side of the carrot-and-stick blackmail and bribery by Russia was that Trump was promised a bribe of a share of Rosneft coincidentally the same size that was just sold off, on credit, to buyers concealed by various cutouts.
Gonna need more hard evidence before being concerned about this.
Occam's razor, but I'm guessing that was autocorrect.
Anyway, no, because there is no incentive for him to do what you are suggesting.
If however you believe his goal is to destabilise U.S. politics, then there is large incentive and zero downside for him to fake a sale of 19% of the company.
When did the people pushing the N Clinton conspiracies every say "this may or not be true..."???
So, your statement is disingenuous, at best.
Fallacy of gray. Just because you can't ask for perfect doesn't mean you can't ask for better.
I've addressed this elsewhere, but there is definitely cases where this can be done.
No, Tom Cruise is not moving to Buffalo, Missouri[1].
No, there is no attempt by lawmakers to legislate the disposal of unused semen[2].
Etc, etc. Some of these are satire, but some are just complete deliberate fakes.
This is a different class of problem to "OMG Hillary is a Nazi/OMG Trump bought Pizza" opinion spam.
"The White House has announced today at 7h30, through speaker ..., a piece of legislation that, quote: ~first paragraph summary of the legislation~.
You can read the full text here.
On the opinion of Bla bla, this legislation is innefective etc etc.
Disclaimer: Bla bla is not related to any of the mentioned organizations, though he declares he tends to agree with the opinion of former president Obama on most issues. You can read more about this commenter here."
According to Rudy Giuliani, when Trump first announced it, he called it a Muslim ban, so...
https://www.democracynow.org/2017/1/30/headlines/trump_favor...
So you admit you are only reading from sources with a similar bias/narrative on this issue?
How is that any different than people in the Bible Belt who believe the earth was made in 7 days ? Or elderly people who become vulnerable and fall prey to exploitation (because they believe phone scams, or sob stories from strangers, etc) ? (Edit: or young people who think they're going to grow up and be an astronaut or NBA star). There will always be a sizable proportion of the population who believes wacky stories.
I know you weren't implying this, but NYT and WaPo's hands aren't clean either. I really didn't appreciate how they pushed the "Russia HACKED the election" narrative. That's predatory on the same level: take advantage of the nebulous word "hacking" and misapply it to something that doesn't even rely on computers, do some hand-waiving, and voila, there's some "fake news" to dish up to the masses [which is ironic when your other stories are about the spread of "fake news"].
I see no difference there but I'm not sure that I get the point.
> I really didn't appreciate how they pushed the "Russia HACKED the election" narrative.
I never got the impression that those news orgs were saying that actual voting counts were being hacked. But the emails were hacked and leaked.
They jumped the gun on blaming Russia, but that wasn't fake news (provided you believe prominent members of US intelligence).
Even Trump seemed to think it was plausible... At least plausible enough to publicly ask Russia to hack Hilary's emails [1].
The NYT and WaPo (among others) are incredibly biased, and much of the problem is that their support base is too sheltered to understand how. I've said more on this in previous comments.
I followed the election from a variety of sources (WaPo and NYT included) and never got that impression. But I agree that it was a circus from all angles.
https://www.washingtonpost.com/news/rampage/wp/2016/12/28/am...
Do you think it's mostly from misleading mainstream news stories? Or from smaller outlets (social media, celebrities, memes, etc)?
my point is that there have always legions of people who believe ridiculous things. They've been around for every single previous election. So we shouldn't be surprised when the scientific community says "civilization has changed the climate" but people still don't believe it (even when the claim is supported by evidence).
I'm sure one of these anti science boogeymen you've created would think you are equally brainwashed and they are equally correct in their views...
I still like the Times to an extent, but they like many other papers get it wrong from time to time. Doesn't mean I should leave and never look back, but instead I have taken to critically thinking about every story I read from every source to come to a conclusion about an issue.
I don't think fake news or poor journalism is a problem. The problem is people and their lack of critical thinking, often substituted for blind acceptance of a narrative that fits their preferred reality.
Fake news on the other hand are literally made up lies. The overwhelming majority of fake news articles have no basis in reality, period. They tend to be written by people who want nothing other than to get page views.
After all, telling outright lies is just amateurish; the NYT has been shaping society's perspective on facts for over a century.
That's right, because when Wikileaks revealed high levels of collusion between media outlets and the DNC/Clinton campaign, placing and pulling articles/stories to present a specific narrative, it wasn't out of desire to mislead, rather it was only a desire to 'correct the record'. /s
America and the world have a situation where advertisers need to prey on the Id and emotions of users in order to sell widgets. The advent of the internet crushed the ability of media firms to stay afloat and replaced the ecosystem with many upstarts who also faced the same pressure.
In short there's an overheating attention grabbing system when it comes to the media, which is its own issue that needs to be solved.
Unfortunately there's a new hitherto unbelievable problem, where people fabricate complete and utter lies/fiction and can sail it down the streams of social media.
This is different from the older issue in that these are not major news agencies, but websites masquerading/camouflaging themselves as a news outlet and fabricating stories designed to directly press emotional buttons without having to deal with facts.
A closer model is conspiracy theories.
I agree this is a problem. It also happened rather equally on both sides.
If you really want to know how it works, please take a look at a recent a paper co-authored by Alessandro Bessi. It came out a week ago. It should be accessible for most people on HN.
I think if we compare today's new to a decade ago's, the change is not more capitalism (ie, for-profit, corporate, quarterly-report focused...). The change is more choice, more information, more volume and the chaotic, many-to-many distribution systems of the social internet.
The clickbait imperative has got a quasi-capitalist single mindedness to it, but I think we'll be chasing down the wrong culprit by characterising the problem as a capitalist one.
In fact, I think it's best if we try not to jump to conclusions at all. We're hitting some problems with the way our democracies work and I haven't really heard a narrative yet that feels satisfying. Lets not jump to conclusions. We need to be right about it.
Is a great justification for doing nothing, ever.
It feels like we are entering a very much unprecedented era in America, where the President freezes out news organisations that do not report on him favorably, and is actively attacking the media with the aim of establishing the official government narrative as the only one voters should listen to.
This is extremely dangerous. I'm not saying that we should rush to judgement, but we shouldn't just sit back and wait to see how it all unfolds either. This is an issue to be investigated and worked on, urgently.
A fake news detector isn't going to do anything by itself to solve the problem. Fake news consumers are going to guzzle the junk regardless of warning tags, scores, or nutritional labeling.
In the hands of a skilled journalist/activist/thought-leader whose abilities are "augmented" by this tool, however, it could really be a powerful antidote.
> All I know is the people who seem prone to believing the stupidest stories
It goes both ways I think. Consider the possibility that there are more people that believe that other 'stupid' people believe these stories than there are 'stupid' people who actually believe these stories.
You can very easily overwhelm the signal with noise. The plethora of independents means nothing if only a few of them get heard.
Matter of fact, with some pushing, you can have one "independent" dominate a large portion of the channel. At some juncture the network effect takes over and the process becomes self sustaining, you just need to play on people's/your target markets biases well.
This is a people problem, and people are getting way too good and understanding and manipulating the biases of other people.
Really, because not more than a few weeks ago I remember a whole bunch of people believing the "we can't verify, but we're going to press anyway" stories about Trump and golden showers in Russia, not to mention a whole bunch of other similar fake news regarding Trump and Trump supporters.
I think you (and many others) are selectively applying a bias to what is and isn't fake news and the type of people who believe it based on your own political preferences.
And therein lies the real problem.
1) media agencies on all sides had the dossier and didn't publish it since t was unverifiable
2) when buzfeed or whomever linked the dossier, it forced other firms to do so as well or lose out on a critical story.
3) the adaptation to deal with this is to not report the dossier as fact, but as claims. And of course, it spiraled downhill from there,
In this scenario it's important to note that media firms avoided reporting on this issue because it was crap.
For global warming, ah man where do we start.
Firstly, it's unfair to lambast people for believing it's a hoax.
Since the 80s and before, a long, intelligent and concerted effort to spread FUD has been active in America. It is today successful enough to spread around the world.
If people were old enough, they would remember that initially scientists would not give the time of day to global warming conspiracist/deniers.
It was correctly believed that doing so actually gave cranks and loons the patina of credibility, when they should be allowed to die out.
Unfortunately, these cranks and loons were cultivated and nurtured, and given airtime. In particular there is one news network that I recall had a more active role in this and it was not CNN.
Eventually scientists were forced to engage, and by engaging they lost.
Scientists, (and people on HN) are constantly making the error that this is about facts.
It's about entertainment and playing with human emotions.
Scientists walked in expecting a debate, when it was about showmanship, with one side reporting things unevenly.
As a result there are people today who will be anti global warming because they are being manipulated.
Fake news is a new problem
Global warming wasn't part of my argument. That was just me quoting the parent who appears to believe only one side of the political spectrum is susceptible to fake news.
Personally I trust the science, and think that even if you don't there are enough other compelling reasons to move on to renewable energy sources that we'd be stupid not to anyway.
WaitButWhy.com had a good series on Elon Musk that explored many of these issues: http://waitbutwhy.com/2015/06/how-tesla-will-change-your-lif...
> Fake news is a new problem
There was plenty of science going against 'smoking causes cancer'. Fake news has been around for quite some time, just not in its current more virulent form.
The day the printing press was invented was the day print journalism was subverted to push agendas.
News is a product, but it should be noted that fraud is a crime in a free market. Not that it discounts the rest of what you're saying.
However, news as a public service has standards of liability to the public in general, not just to your own customers. It is an obligation to society, not to the buyer.
Capitalism isn't going anywhere obviously and shouldn't. On the other end is state-run media, e.g. North Korea. We wouldn't want things to swing in that direction or even toward RT or many other examples.
I think we should be looking to replicate whatever magic is and has been in PBS, early BBC, ITN, and other earlier UK and French programming.
You can see this right now. The mainstream media falsifies or omits the truth on behalf of their backers or the organizations they fear: major political parties, subsidiaries of their parent companies, major investors, etc. I would argue that they do this because the public is not their customer, it's their target demographic.
I don't see the problem. I don't want to hear lies, therefore I create a demand for true news. I don't need a nanny facebook/google/state telling me what news I like, and neither does the rest of the population.
Frankly, I find it disturbing that so many well educated people aren't able to objectively think about an actually objective problem.
The fake news that started the concept of "fake news" is not a subjective problem. The problem is literal invention of facts not even related to reality, combined with the mass distribution of those invented facts. It is a problem of mass deception.
The classic example: Millions of people shared a post saying the Pope endorsed Donald Trump.
This is objectively false. It's not about liberal vs. conservative. It's not about whether it fits your worldview or not. It simply isn't true.
During the election there were so many literally false stories that got a very large amount of attention.
That is the problem. And yet, so many people seem to think it is a political issue.
It's a sad world we live in where even objective facts no longer matter.
I think that is why so many people are terrified of the direction things are headed.
That's a hyperbole -- most of the dissidents here do not deny that completely fake stories occur, but express valid concerns about what kind of things will be labeled as facts or not. Based on what we've already seen (the WaPo list, for example), there are publications they would like to censor with very good track records, including the "Ron Paul Liberty Report." Production budget aside, that's often the first source that reports on interesting new bills submitted to the House, such as the recent pre-authorization for invasion of Iran bill. This is not a fake fact -- you can see it on the House website. And there aren't supporting examples of fakeness provided by WaPo for their list, even though I could find numerous factual errors by WaPo or NYT, like the power grid hacking or Aleppo (usually fixed within a few days, but still).
And of course, there's the fact that powerful propaganda does not need to stem from provably false facts -- it uses selection/omission of stories and crafting heavily-opinionated narratives to 'mislead.' That is just as concerning as fake facts, and a poor attempt to stem the latter would most likely end up stifling unpopular opinions and further empower the propaganda you see on major sources like WaPo/FoxNews.
Same with xenophobia: a xenophobe doesn't marry an immigrant.
Same with misogyny: a misogynist does not appoint a woman to run his campaign.
Same with antisemitism: his daughter converted to Judaism to marry a Jew, who also had a hand in running his campaign.
Same with racism: he wasn't in any way called "racist" by anybody until he ran for president, and there are numerous photos of him with prominent members of the black community dating back decades.
Much of the slander we've seen about the guy is totally fake. It seems HRC had very little of substance other than the pussy grab tape (note that there's no evidence it was anything other than talk), so they just made shit up and told CNN to air it 24x7. And CNN gladly obliged.
When the same people that manufactured and spread fake news start lecturing us about the harm of fake news, I just facepalm.
This is exactly what I'm talking about.
The reason we're talking about fake news is NOT because of the traditional mainstream media. It's not because of left wing websites such as Alternet or extreme websites such as Breitbart.
The reason we're talking about fake news –– and the real problem that absolutely has to be addressed –– has nothing to do with those websites, or with nearly any media organization that has an easily identifiable group of people running it.
Sure, those websites might post factually incorrect information on occasion. But that's not why we're talking about fake news.
We're talking about fake news because there were –– and are –– websites that repeatedly posted blatantly false "news", during the election, and got massive distribution on social media networks such as Facebook.
That is a problem that has nothing to do with your opinion of Donald Trump. It is not clear why you brought him up, except to distract from the issue.
I'm just pointing out that the "legitimate" news sources aren't that much better than Breitbart these days, except a lot of people treat them as though they are.
I am willing to be that people called him racist when the Justice Department filed a housing discrimination lawsuit against his company in 1973. Or when his company settled that suit but in 1978 the government filed another complaint claiming his company didn't meet the terms of the settlement. https://www.salon.com/2011/04/28/donald_trump_discrimination...
This comment also shows another problem that is harder to solve than "fake news," which is the interpretation of facts. What you call just talking, I call bragging about sexual assault and even if it didn't happen is something that I would call someone a misogynist for bragging about in the first place. Other people don't.
This, in a sense, is Donald Trump's fault. He is so good at controlling the conversation; his bombastic style demands a response; this is the same style you're using in your posts, and it has successfully redirected the discussion. It doesn't matter whether you're right; the topic has now changed.
This doesn't get things back on topic. Though, it took minimal effort to find instances of Salon running ridiculous reports. [0]
Here the real problem: these companies can't get their hands on content fast enough. WaPo produces 500 articles/day and 1200 items/day. AFAIK, because there's no real punishment for quasi-spurious claims, there's no incentive to spend marginal time vetting sources. Today, the story can be shared 100K times in the time a retraction takes place. But by then, the damage has been done, and again, the punishment isn't material. I don't think a retraction fully undoes the damage of spurious reporting.
[0] http://www.slate.com/blogs/the_slatest/2013/01/23/salon_trut...
Hence the problem. It is, it's not rocket science this is true.
But AI or not you seem like you won't change your mind. This is what the right thinks about Fox.
I gave evidence from a 2011 article that has a number of links describing how he was involved in things in the 70s that would make someone call him a racist.
It's prejudice against a class, and it's extraordinarily common for otherwise small-minded people to befriend or get close to (and defend!) individual members of that class. That does not actually make them less racist, misogynistic, homophobic, etc. It just makes them a hypocrite.
"I can't be racist I have a black friend" is basically what you're saying, writ large.
Pretty sure people were calling him a racist when he was running around promoting the Birther hoax - I know I was - or even before that when he called for the death penalty for the falsely accused Central Park 5 in 1989.
But fake news is not the same and it's a whole different issue which also needs to be resolved.
Trump was first sued by the government for racial discrimination something like 40 years ago; he's been being called a racist for a long time before he ran for President.
Of course, a lot more people paid attention to him and talked about him at all, and a lot more people paid attention to what was said about him, once he became a Presidential candidate.
Obviously, the national attention on people doing so was less before he ran for President, but the idea that he only was accused of racism due to his Presidential run is false.
They will not, and the reason has to do with language. Ludwig Wittgenstein tackled this 100 years ago. The best that machines will do for you is to label something as true or not as if you had consumed the article and decided on your own
That is a completely different thing from identifying fake news or truth
There's some value here. There are also some hard stops. You'll find them :) Best of luck, guys!
A Google Image Search of the picture in the fake article would have found the picture is much older, and isn't what the caption described.
If I'm tech-savvy and skeptical, I might have done that when consuming the article, and decided on my own it was fake. Machines can do that automatically for people that aren't tech-savvy, or that don't start reading it with a skeptical mindset.
There are facts of course - like such and such event happened or not, such and such statistical data is showing this number of not. But it is never served raw, it is always garnished with opinion. E.g. the fact is that Trump signed certain executive order. But you can say it's "Muslim ban" or "temporary restriction on admittance of nationals from terror-ridden countries". Not the same.
And various "factchecking" sites are fully complicit in this and inject opinion as often as not - to the point exactly the same thing said by two different people can be deemed "mostly true" or "mostly false", moreover, exactly the same thing said by the same person can be "mostly true" and "false".
You can't hope to have robots make sense of it if we can't. You can, of course, build a neural network and train it, but it would just put "robot approved" stamp on biases of whoever trained it.
We need more cynical takes on things.
It's utterly appalling that on one side we have right-wing extremists yelling "FAKE NEWS" about CNN because CNN might get something wrong one time in ten, implying it's no different than Breitbart or Fox News which are wrong nine times in ten, and on the other we have left-leaning people beating up CNN for making mistakes and suggesting their trust is completely eroded.
News organizations offer a view. You should be prepared to vet everything you read from anyone. We need to give people a toolkit to help verify stories, to develop their bullshit instinct.
"re-transmit" sounds neutral but "re-re-transmit" sounds passive aggressive. This would be very hard for a machine to interpret
Maybe news could rely on first principles approach. Quantify things like political stability, information availability, cultural problems, etc. and maybe apply a first principles approach and maybe an algorithm could tell use what the hell is going on at any given moment and maybe give us some insight into what may go on in the future as well.
I would assume intelligence agencies have something similar to the above. The problem is things happen in realtime and if a regime falls or a state fails or whatever, we have to adjust our models immediately.
Throwing out hands up and saying Wittgenstein solved it seems lazy but idk.
Good luck. It's a worthy goal, but we are very far away from being able to do this in a rigorous fashion. There is a reason academia has the distinctions of "hard science", "soft science", and "not science". Cultural and political dynamics, by their inherent nature, might never be objectively quantified.
I imagine if we had this tool when George W invaded Iraq, this AI would have said Weapons Existed, being all the CIA, and Politicians said so.
Isn't that how your filter works? Anti alternative media, pro establishment?
This comes up often and is untrue. The Intelligence Community did not have confidence in the WMD claim.
Good read regarding this: http://www.newyorker.com/magazine/2003/05/12/selective-intel...
As I said in another post: The only thing you can possibly check for with any reliability is the validity of the base, source event, if there is one. Any higher derivation in the wrong hands (Facebook, for example) is systemic censorship. How can we safely do that? Recursive link tracking, perhaps, to a "primary source," which relies on the AI matching the factual claim to an appropriate source, either a video, image, or text that can be considered the legitimate source based on context (for example, an article about Apple's earnings could link to a primary source of their press release).
Regardless, the most dangerous form of "fake news" is by the selection/omission of stories, and that seems pretty impossible to quantify with today's technology (and of course the most prevalent on all of our 'real news' sources).
This. Very true, and woefully absent from most discussions regarding "fake news".
An algorithm could watch, say, 50 sources and see if they simultaneously start using identical phrases, given access to a thesaurus aren't those odds a little unusual? This happened a couple times in the last election. You'd have a much simpler task of quote analysis of course, to exclude directly quoted content.
Then factor in something to detect press releases and video news releases where the verbiage of large sections is identical, this strategy is almost too easy.
Another factor is "large known propaganda sources" simultaneously releasing the story but small propaganda sources don't operate at all because fundamentally there is a size cutoff. The Wash Post will be in every propaganda operation because they are huge, but the "nowheresville register" certainly will not. Real news would of course be covered by both, so it should be very easy to detect organic self organizing behavior vs small scale conspiratorial behavior.
Now algo-proof propaganda could emerge much like handwritten letters to your congressional representative, before being inevitably ignored, are considered more impressive than mere form letters or signing a list. But it would detect at least some fake news based on fake news being part of a coordinated centralized mass produced campaign rather than trying to detect fake news based on its inherent fake-ness itself. Sorta like swatting bugs that bite you or try to, rather than identifying each individual bug then swatting only the mosquitos. Its very difficult to detect if something is a Walmart product, but its easy to detect if its mass produced, if the quality is very low, if it was made in China, if it contains chemicals banned in the USA for safety reasons, and in the long run works just as well.
What would that mean? Let's say I write an article where I claim White House is controlled by reptiloids from planet Nibiru. Would it check that I have any links in my article at all? Or that the links actually prove my claims? Or that these proofs can be trusted (I may just link to my own website or one created by my friend)? I could insert a hundred links to a most respectable austronomy sites to prove that planet Nibiru exists, but how would you know whether these links actually prove it? I could refer to a lot of actual news content as the proof White House occupants can't actually be human. How you evaluate if these links actually support my claim?
It's more difficult than just blindly blocking any stories that aren't covered on 'major, trusted sources', and probably only a couple of organizations could implement anything close to this right now, but it's the best we can strive for. All the other 'solutions' I've heard involve censorship of unpopular opinions.
Edit: But yes, to the point, my envisioned implementation would not find any referenced sources in your article or through a search, and if it did, those sources would not be authorized to make the claims about the Whitehouse reptiles, so it would get the notation that indicates 'could not verify.'
At first, most real news stories would be unable to be verified, but over the time, there would be fewer false negatives. The important thing to recognize, though, is that the notation just means the reader has to make his own judgment; it does not state whether or not it's fake.
And "real stories" should also probably have a notation with a direct link to what the algorithm thinks is the base source.
1. Encouraging journalists to actually link to a source at all. Because at the moment, a depressing number of them don't even bother sourcing anything. So you end up just having to 'trust' the organisation that the story has been checked out beforehand.
It's an especially big problem with web versions of printed newspaper articles, and is a general problem with print newspapers as a whole (since due to space limits, they don't publically source/cite anything).
2. Getting journalists to link to the original source. Because at the moment, you'll see news sites sourcing other news sites which then source other news like a lengthy game of Chinese Whispers or Telephone. It's how a lot of 'fake news' makes its way in the media food chain.
So yeah, something like this would be useful just to get publications to cite their sources better. Or to make them look for the real source rather than one of a million middlemen.
And of course nobody would love this technology more than the Chinese and Russian governments. Is a system aimed at quickly identifying obscure blogs that disagree with "high-credibility" sources supposed to help democracy?
It's one of the reasons why our employment choices and our decisions around who to work with must be taken in the context the impact it will cause on the world around us. As we begin to automate more and more of traditionally human jobs, we're going to find our work inextricable from the political landscape our systems are creating.
Friederich Nietzsche
Q: Why did you choose the stance detection task rather than the task of labeling a claim, headline or story True/False, which seems to be what the fake news problem is all about?
A: (...) Our extensive discussions with journalists and fact checkers made it clear both how difficult "truth labeling" of claims really is, and how they'd rather have reliable semi-automated tool to help them in do their job better rather than fully-automated system whose performance will inevitably fall far short of 100% accuracy. (...)
Q: OK, but what does stance detection have to do with detecting fake news?
A: (...) From our discussions with real-life fact checkers, we realized that gathering the relevant background information about a claim or news story, including all sides of the issue, is a critical initial step in a human fact checker's job. One goal of the Fake News Challenge is to push the state-of-the-art in assisting human fact checkers, by helping them quickly gather the information they need to make their assessment.
In particular, a good Stance Detection solution would allow a human fact checker to enter a claim or headline and instantly retrieve the top articles that agree, disagree or discuss the claim/headline in question. They could then look at the arguments for and against the claim, and use their human judgment and reasoning skills to assess the validity of the claim in question. Such a tool would enable human fact checkers to be fast and effective. (...)
This means they're very much aware that 'solving the fake news issue' in a fully-automated way is a folly, so they are instead looking for tools to classify and retrieve corroborating or dissenting reports about the same topic. I feel this approach demonstrates an awareness of the problem, addresses some of the criticisms raised in this thread, and could lead to useful tools and datasets down the road.
They've been beaten at their own game by an opponent better at using modern/popular methods of disseminating information.
Oh right, because we can all agree on which news organizations are credible. /s
https://www.washingtonpost.com/lifestyle/style/its-time-to-r...
Also somehow CNN shows in the front page of result for "fake news" image search on Google. It wouldn't be totally wrong, depending on how the term is defined.
Heh, I think need an explicit meta-fake news category.
Does anyone know if this challenge has anything to do with Media Matters / Shareblue?
If you scroll to the "About" section you can see who's who and do some digging to see if connected.
http://www.fakenewschallenge.org#about
Or just contact them and ask (info@fakenewschallenge.org).
"Media Matters has already secured access to raw data from Facebook, Twitter...
Predictive Technology
Bringing this data analysis to scale will also allow Media Matters to identify which individuals and outlets are the most destructive forces driving fake news, misinformation and harrasment.
Cutting-edge advances in cloud computing and machine learning will enable us to identify patterns and connections that would otherwise go under the radar. ... Media Matters' core budget for 2017 is $13.4 millions which covers a staff of 81. It is a 501(c)(3) nonprofit.
This budget allows us to create a 34 person research deparment engaged in media monitoring, research, deep dive analysis and rapid response... this budget also alows us to invest in technological innovations like the creation of an early warning system to identify in the profilerationof fake nws more efficiently and to create cutting-edge predictive technology that allows us to identify patterns and connections in order to asses how misinformation will move and how we can neutralise it. "
Quoting Wikipedia policy itself:
>If Wikipedia had been available around the sixth century B.C., it would have reported the view that the Earth is flat as a fact and without qualification. And it would have reported the views of Eratosthenes (who correctly determined the earth's circumference in 240BC) either as controversial, or a fringe view. Similarly if available in Galileo's time, it would have reported the view that the sun goes round the earth as a fact, and if Galileo had been a Vicipaedia editor, his view would have been rejected as 'originale investigationis'.
A few comments based on what I've seen here so far:
"Fake News" might not be "the problem", but it is a problem. There is a real set of completely fake stories with no supporting evidence. Taking this away from politics for a minute, the archetype of this is "celebrities XX is moving to town YY"[1]
Comments around "this is just mainstream news attempting to justify its role" absolutely and completely have a point. There is plenty of room for new players to fix this problem (and hopefully be rewarded for it), on both sides of the political spectrum.
Comments around "The NYT does fake news - see their Iraq war advocacy" are somewhat misguided IMHO. The NYT has a number of problems, but no one should make the mistake of confusing opinion, analysis and forecasts for news. In the NYT Iraq War case, their analysis and forecasts were completely wrong and their opinion was misguided because of that. It is completely fair to make the point that their reputation suffered because of this, but we need to separate that from news reporting. This is something that news organizations don't do very well at the moment, IMHO.
As has been pointed out, this won't fix the whole problem. Most of the "Yes, but..." discussions here have been had when trying to come up with a reasonable but helpful task. I think stance detection is a reasonable subset of the problem.
That said, I think the other part of this problem is how the news is consumed. Our lazy minds are trained to read the shorter version if there is one available. Every day more people are liking the 140 character version of news. The more you read (from more people), the more you're reinforced in that belief. So while we train the models on news articles, we should also attempt to train on tweets. Just my two cents.
The first step to identifying 'fake news' (even in NYT or WSJ!) is to be able to pull out attributions from the source text.
This would give you some idea of where the information in an article actually came from. "Fake news" will tend to be either unattributed or misrepresent the thing it's quoting.
But a good first step toward better media literacy would just be giving people a representation of who actually said the the things in an article, and that seems much easier.
So who decides what organisations are credible? Or orders them by their credibility? Someone has to create the list. Someone with obvious human biases towards one view or another.
Even if all the platform did was accept "news stories" and transparently verify it's authenticity I think people would clamor for it.
Of course the devil is in the details because that would require proper investigative journalism, which may be prohibitively expensive.
Fake news is a meta problem, and it's solve by funding news agencies and removing them from the advertising loop, as well as laws which make news vs opinion different discussions.
Tech cannot solve it without also creating tools which will be used against us.
The magic tech solution to solve fake news is essentially an assessment of the intent of the author.
Any tool which even gets close to ascertaining this will be badly misused.
function isFakeNews(news_piece){
var list_of_legit_sources = _get_current_approved_sources(); // makes synchronous http request to illuminati.com/api
if list_of_legit_sources.inArray(getSource(news_piece)) return false; else return true;
}
After that, you can build an engine to crawl news articles or have them submitted, and detect the claims being made. Those claims can be either auto-matched to existing ones or create new ones (deduplication) and then the site has a fresh new stream of claims to fact check.
StackOverflow could build this site on their existing engine. If someone here knows them, can you reply and put me in touch?
Where the facts lie in many subjects is not immediately obvious.
Algorithms like this are a good idea provided that they're used wisely. I expect them to detect conflicts between treatments (with false + and false -), which would be useful to humans making their own judgements. Inflicting their results, without a chance of additional vetting, would not be a good idea.
And I'm not suggesting all news agencies [whatever this means in 2017] should transform into NPR or PBS clones. I like MSNBC, FOX, CNN , ABC... news but they all exist within media empires. Let's figure out how to spin them out into entities not funded from media business models. Hard problem, sure. So what. We have to figure this out.
I see what you did there ;)
In all seriousness, though, if people do not learn to read the news and media critically then an AI system isn't going to make things better. I remember back in high school, here in Canada, being taught media literacy. I think it was because of our close proximity to the US, and how influential US media can be.
This is not a new thing, and we used to simply ignore it. Now we made it a thing.
Exactly. I'd argue that disingenuous news from mainstream news organisations is worse than dedicated "fake news" websites because they exploit legitimacy and reader-trust to peddle falsehoods. They are our port in a storm of bullsh*t, and they're failing us.
In fact, I would go as far as saying that the internet, itself, was designed and built in the naive belief that humans would better themselves by instant access to knowledge. This past decade and a half has completely shattered our 1980s and '90s vision of the promise of the internet.
We have few means of combating propaganda that threatens our values, reclaiming our privacy, or maintaining our remaining sanity. When a shared objective reality no longer exists, we're all loons.
Consider the sensationalist headline "Trump Facing Impeachment for Muslim Ban".
Okay, a lot of people (including politicians) are saying "yeah we should maybe impeach that guy", and a lot of Muslims are affected by the order. But there's no impeachment trial and it's not a direct ban on all Muslims.
I don't think it's appropriate to shift all the blame to the publishers but, when people just read the title, nod approvingly, and close their browser, it's the same effect.
When we start with something substantive like this and take it straight up to generic ideological soundbites, we cheat ourselves of the kind of intellectually interesting discussion that we might've had instead.
How do you debate a person that denies objective reality, and embraces their very own alternative reality?
I don't think you can. For the same reason you can't negotiate with terrorists. They aren't playing by the same rule book. You hold out a hand to shake, they stab you in the face with a knife.
Far too many people are unwilling to recognize that news that confirms their world view might be fact. They cannot or will not challenge it, and simply accept it as fact.
> prioritize refugee claims made by individuals on the basis of religious-based persecution, provided that the religion of the individual is a minority religion in the individual's country of nationality.
You're right. It's not an order that bans Muslims, explicitly. It's an order that bans them implicitly. The NPR annotation:
> If you CTRL+F this document, you won’t find the words “Christian” or “Muslim.” But if you read closely, this is the section that indicates prioritizing Christians. How so? Because the seven countries affected by this order are all Muslim-majority countries. Also note that Trump himself indicated a preference for prioritizing Christians in an interview with the Christian Broadcasting Network. And former Trump campaign surrogate Rudy Giuliani told Fox that Trump first called it a “Muslim ban” but asked him to come up with a legal framework for it. Giuliani said he did, one based on “danger,” not religion.
He did ban Muslims, just not all of them (yet.) And he did talk about banning all Muslims on the campaign.
Now the Kenyan thing requires to believe that a lot of people and countries are on it, and have been decades before Obama was anyone of note.
When Obama bombed those very same countries was he bombing Muslims? Isn't that genocide?
Only on Google cache since he removed it:
"December 07, 2015 - Donald J. Trump Statement on Preventing Muslim Immigration
(New York, NY) December 7th, 2015, -- Donald J. Trump is calling for a total and complete shutdown of Muslims entering the United States until our country's representatives can figure out what is going on. According to Pew Research, among others, there is great hatred towards Americans by large segments of the Muslim population. Most recently, a poll from the Center for Security Policy released data showing "25% of those polled agreed that violence against Americans here in the United States is justified as a part of the global jihad" and 51% of those polled, "agreed that Muslims in America should have the choice of being governed according to Shariah." Shariah authorizes such atrocities as murder against non-believers who won't convert, beheadings and more unthinkable acts that pose great harm to Americans, especially women.
Mr. Trump stated, "Without looking at the various polling data, it is obvious to anybody the hatred is beyond comprehension. Where this hatred comes from and why we will have to determine. Until we are able to determine and understand this problem and the dangerous threat it poses, our country cannot be the victims of horrendous attacks by people that believe only in Jihad, and have no sense of reason or respect for human life. If I win the election for President, we are going to Make America Great Again." - Donald J. Trump"
"December 07, 2015 - Donald J. Trump Statement on Preventing Muslim Immigration (New York, NY) December 7th, 2015, -- Donald J. Trump is calling for a total and complete shutdown of Muslims entering the United States until our country's representatives can figure out what is going on. According to Pew Research, among others, there is great hatred towards Americans by large segments of the Muslim population."
It may be gone from the location you were looking at, as happens with all kinds of stuff on the web. I believe this is the same document, available from the DJT campaign site.
https://www.donaldjtrump.com/press-releases/donald-j.-trump-...
The reason I don't believe the "concern" coming from Democrats is if they actually cared about the Syrian people they would have been protesting in capitol hill while Obama was dropping bombs on their villages and hospitals. I was there, the Quakers were there, even some Republicans were there. Sadly the Democrats were not to be found in any significant number any of the dozen or so times I visted.
"Fascists believe that liberal democracy is obsolete, and they regard the complete mobilization of society under a totalitarian one-party state as necessary to prepare a nation for armed conflict and to respond effectively to economic difficulties. Such a state is led by a strong leader—such as a dictator and a martial government composed of the members of the governing fascist party—to forge national unity and maintain a stable and orderly society. Fascism rejects assertions that violence is automatically negative in nature, and views political violence, war, and imperialism as means that can achieve national rejuvenation. Fascists advocate a mixed economy, with the principal goal of achieving autarky through protectionist and interventionist economic policies."
When I read that I see elements of fascism in many governments around the world, with very few of them coming close to the definition.
Even so, you could fairly easily set up a number of points to check off on and score a government (and by extension) it's supporters on a scale from 'definitely not fascist' to 'definitely fascist'.
The question then becomes at what point on that scale it becomes reasonable to call a government fascist and to call a support of that government a fascist. I'm pretty sure we're not there yet but I am also pretty sure that we are slowly but surely sliding in that direction.
Using the term fascism is like a shot in the arm, it tries to wake up the organism that it is directed at that it is getting ill and needs to fend off some viral element. Unfortunately usually the opposite happens and instead the recipient will wall themselves off from the provider of the message. Even so, I guess that even fascism light is dangerous and ugly enough that plenty of people would not be caught dead being labeled a fascist, especially not when there is some hard evidence that it is so.
Probably because it often feels like that's the intent of people name-calling others "fascists."
The best quote about fascism is and remains: "In Italy, fascists divide themselves into two categories: fascists and antifascists."
(background: Russia might have "kompromat" on President Trump...blah blah )
Headline: President Trump appoints "friend of Russia award" winner as Secretary of State.
Just put a "Alert" or "Fake news" button. A lot of political writing can be considered "fake news" depending on our biases.
Obviously followers of these sites will fight back against any "fake" labels, by arguing there is a "cover-up" by organized opposition, that their truth is being suppressed, that any fake news classifier intrinsically has a liberal bias, and so on. The difficulty for me is not identifying what's fake, but convincing anyone else that it is.
Ironically this may have done Clinton in with people thinking well, if she's won already then I don't need to go and vote.
Hindsight is always 20/20, it is very easy to see Clinton winning as a 'narrative' pushed with an agenda whereas based on the data available to them that simply made good sense.
See also: Trump expecting fully to lose (in his own words at some point).
With the issue of opinions derived from factual events, which fit the bill of being 'misleading' (to some groups, at least) -- how do you draw the line? The NYT is certainly not exempt from that (see: half of Paul Krugman's articles).
The only thing you can possibly check for with any reliability is the validity of the base, source event, if there is one. Any higher derivation in the wrong hands (Facebook, for example) is systemic censorship. How can we safely do that? Recursive link tracking, perhaps, to a "primary source," which relies on the AI matching the factual claim to an appropriate source, either a video, image, or text that can be considered the legitimate source based on context (for example, an article about Apple's earnings could link to a primary source of their press release).
All of this would be much easier if there was some "standard" for writing news, sectioning off opinions that derive 'non-obvious' conclusions, linking primary sources, etc. But there isn't, and you would have to censor those who did not conform if there was one, which is not a good solution.
Mainstream news orgs tend to be better about including citations, and sources. They are also better about publishing retractions, when they are caught saying something factually wrong.
I'm as anti-Trump as it gets, but the idea that NYT or WSJ are bastions of reputable journalism is ridiculous.
See, for example, the list of controversies associated with the output of Pulitzer prize winner Thomas Friedman.