"This problem isn't that hard. You filter this stuff and rank it by its SOURCE, not by the content of individual articles."
I don't think reifying ad hominem into code is the solution to the problem.
Of course, if you don't mind a rare few false positives here and there on articles, I've got your filter right here:
def source_is_trustworthy(source):
return False
Remember as you sit here thinking through your exceptions what we're talking about; if your "trustworthy" source was confidently telling you about how Clinton was going to win, it's not one of the exceptions. I won't say I was confident Trump would win, but I was certainly less suprised than most; the fact that I generated this somewhat more predictive model by tossing out pretty much every "mainstream" news source is not a good sign for them. And to be honest, Trump news isn't the only thing that I find this useful for. Beyond the bare facts, most media outlets really aren't good for much anymore, and do you ever have to de-spin their news just to get those bare facts in the first place. (I can't, however, pack up into a recipe how to do this yourself. I think we're in a period of transition in the news industry, much greater than just "the internet makes the old dinosaurs stumble" makes it sound, and I'm at the moment not really all that confident in anything.)
As someone who vehemently opposes Trump, I feel that allegations of anti-Trump bias in the mainstream media were entirely correct and somewhat to be expected. Although Trump is certainly a name that sells papers, Trump's repeated threats to open up news media to broader libel laws were not well received. The news organizations' endorsements were pretty one-sided[0]. I don't want to take you too literally, but it seems to me pretty obvious that a model ignoring the MSM entirely would not be preferable to one that took into account the MSM opinions and then applied a correctional factor. I'm sure we agree that truth is a function of our means for determining truth, but I suspect we disagree strongly on the reliability of Internet news sources.
You must have missed the leaked emails in which Clinton exerts massive influence on the media, planning every last details in exchange for all kinds of rewards [0]
I did miss that. Was that article intended to support that view? The incidents listed don't seem to have been terribly well planned or executed, or to have garnered any positive results. That seems difficult to reconcile with the idea of a powerful media conspiracy.
"source_is_trustworthy" applies to all sources, not just "MSM". As I said, I don't have anything right now that I can point to and say "If they say it, I trust it."
I used to try to apply a correctional factor, but I've found it not even worthwhile anymore. I honestly don't know entirely why. I don't know if the news has gotten that much more biased to the point where it almost swamps the signal entirely, or if the combined financial pressures away from expensive real reporting and towards click-bait headlines has removed the signal, or what, but beyond very bare facts they just aren't worth much anymore. It is also possible this has always been the case and I wasn't aware enough to realize it; there's some classic history stories that back that possibility up.
I do know primary sources are getting easier and easier to consult directly. Which could itself also be a reason my opinion of them has plummeted so much in the past 10 years; it was a lot harder to see through media spin and simplification (deliberately for their audience, and accidentally when they in fact don't understand what's going on themselves, see especially science journalism for that) when they were the only source of information.
It won't be reached in practice, but if you make it the policy, you have an argument to take down any bullshit. With 50:50 it's going to be an endless quarrel over who has had enough so far, who is underrepresented and why everything is dominated by two polar opposites with no voice given to people who don't want to belong to either camp.
The case of news sources failing to predict events seems like a red herring in the larger debate here - news media is generally about reporting events that have actually happened. Predicting presidential elections or other uncertain events is a side hustle they've been dragged into because it sells. And if you throw out their predictive failures, I think it's pretty easy to compose a trustworthiness function based on overall factual accuracy. And sure, there may be lingering biases that need to be corrected for, but even so, the MSM biases generally don't extend into the realm of pure fabrication, whereas clickbait outrage farms absolutely do - and they are a large chunk of the problem.
PageRank is also not unbiased, it has been played before many times and pure PageRank would be pretty easy to play - that's why Google keeps tweaking it years after it was invented. And still SEO industry exists. Now, SEOs just compete on being on the first page, not claiming something as grandiose as "truth" - add that to the mix, and all the biases there are, and you get pretty much hopeless task. And of course you get under a constant stream of criticism, and maybe also government regulation - it's one thing to just produce search results (even that is controversial, remember "right to forget"?), it's another thing to claim being able to "detect falsehood" - that's very sure a way to get sued and for the government to get involved.
To determine correctness, the adjudicator requires knowledge outside the system. In the scientific method, this is done by generating falsifiable hypotheses and performing the experiment. Failing that, you have to fall back on heuristics, like "what I already believe", Occam's razor (simplest theory consistent with observations wins), or reliance on authority/consensus.
Consistency is simply majority wins, which completely stifles new ideas or minority opinions.
> Consistency is simply majority wins, which completely stifles new ideas or minority opinions.
No isn't.
You don't need to have a "non objective view of the truth" - i..e you don't need the ability to ask the system "Is this statement true?"
You just need the ability to ask the system "Will I think this is true, given the other things I have claimed to be true?"
Maintaining internal complexity of your own claim graph gets really tricky unless you keep it as accurate as possible.
For example, consensus view was that trump would not win the election. Trump did win the election. Thus, those two claims:
* (before) Trump will not win
* (later) Trump won
These two are inconsistent. It's possible to rectify that inconsistency by marking the earlier statment cancelled. So now we have three statements
* (before) Trump will not win
* (later) Trump won
* (now) I was wrong to say, "Trump will not win."
The gap between "now" and "Before" can be used to compute "probability of a statement being retracted." The smaller those gaps, the more likely it is what whatever you say comes with an asterisk next to it, saying "the expected time to retraction of this statement is: x days"
Ah, you're only speaking of temporal consistency (or "constancy" - does the source's statement change over time).
I was speaking of consistency with the general body of knowledge in the system, (correspondence?).
In the specific case of evaluating major news sources for trust, I think the vast majority of their statements are never retracted/changed, and therefore they would rate highly on an overall "constancy" scale as well. Their many statements about Trump's likelihood of winning were very constant as well - for months they said he'd lose, then they changed, and forever after they will say he won.
Sounds very temporally consistent, with just a single change.
Not sure if the adjudicator could narrow it just to "trust of the source's future predictions about politics", but basically the whole country just got a negative feedback on the trust weighting for major media in that domain.
I don't think reifying ad hominem into code is the solution to the problem.
Of course, if you don't mind a rare few false positives here and there on articles, I've got your filter right here:
Remember as you sit here thinking through your exceptions what we're talking about; if your "trustworthy" source was confidently telling you about how Clinton was going to win, it's not one of the exceptions. I won't say I was confident Trump would win, but I was certainly less suprised than most; the fact that I generated this somewhat more predictive model by tossing out pretty much every "mainstream" news source is not a good sign for them. And to be honest, Trump news isn't the only thing that I find this useful for. Beyond the bare facts, most media outlets really aren't good for much anymore, and do you ever have to de-spin their news just to get those bare facts in the first place. (I can't, however, pack up into a recipe how to do this yourself. I think we're in a period of transition in the news industry, much greater than just "the internet makes the old dinosaurs stumble" makes it sound, and I'm at the moment not really all that confident in anything.)