The Information Apocalypse (2018) | Hacker News Reader

The Information Apocalypse (2018)

diracdeltas.github.io

105 pointskochihabaya54 comments

I generally agree with the article that people get too caught up in imaging sci fi deepfakes incidents swaying entire populations, when the economy of beliefs is way more complex than that.

People have been able to fake photos for a hundred years, and yet fake photos aren't commonly sparking national emergencies. It's not because photos aren't sufficiently lifelike, it's because the ways we validate that information are very straightforward.

Photographer and subjects mysteriously unavailable for follow-up questions? Huh.

Meanwhile, people thrive on validating misinformation.

TFA is right this is a psychological problem not a technical one.

In the 1951 study "They Saw a Game," students supporting each side of a football match counted a different number of infractions and reported referee bias in opposite directions.

Key takeaway: It is impossible to have a neutral arbiter that is also perceived as neutral.

That's why even the most innocuous filtering and sorting algorithm is accused of pushing a bias by interest groups.

If you have the AP cryptographically sign "real" news, the only thing that will do is sentence the AP to allegations of bias.

And this isn't that new. Journalism in the US used to not even bother trying to be objective. Papers would have Republican or Democrat in their name. Political Reddit threads, rabble rousing echo chambers that they are, are also a reversion to the mean.

This may be intractable, but I think our best hope is greater education about cognitive biases, to give people the language to discuss biases with each other that's more clinical and less charged, and to help people recognize the most common traps.

Or maybe just know this is an old problem and just hang on for dear life.

Yeah, my whole impression while reading this article was "why try to solve a non-technological problem using technology?"

If the over abundance of information is just reinforcing existing psychological issues with humans then humans have to either get over these issues or reduce the exposure to information.

How are you going to reach people nowadays, except via technology?

Also, we're pretty much on a timer right now to get things right on at least climate change policy. We don't have the luxury of 30 years, or even a decade, to 'get over these issues'.

You can simultaneously believe that 'all the kinks with new technologies got worked out eventually' and 'it really sucked for the ones who didn't survive the transition'

We cure cancer, which is biological issue, with chimical means, and so coming from technological advance.

That said, that doesn't means we have the technological means to solve problems posed by fake news and radicalization of opinion.

Regulation is probably the key, even if it means policing the web.

The problem here is fake news can comes from everywhere and is not restricted from coming from countries outside the national intended audience where national regulating instance have no influence.

An international watchdog has to be formed.

As for biases... I think in many discussions these could also be called opinions. And you are allowed to have one of course. So I doubt a formalization of language is helpful here at all. Trying to find the minimal common position is probably more constructive and you would need to work out things from there. Conflicts will arise inevitably that need to be dealt with by compromises.

The problem with attesting cognitive bias is just that you could invalidate any statement without proof, since you cannot measure bias. Cognitive bias has few measurable effects with different results from small variances in inputs. Wouldn't want to invent recursive cognitive bias here. I think cognitive psychology itself would need to take a closer look at itself. It can very well be an indicator that formal logic and heuristics are not a sufficent model for human thought processes. Sounds trivial really. Maybe you believe otherwise because you heard someone saying something stupid. Bias?

As for vetting information, there simply isn't any authority. Everyone could become one, you would just need to convince all the others. I also often suffer from the circumstance of being the only one with the correct opinion or bias for that matter.

> students supporting each side of a football match counted a different number of infractions and reported referee bias in opposite directions

Exactly. Did they report their numbers because they thought them to be true?

> As for biases... I think in many discussions these could also be called opinions. And you are allowed to have one of course.

Of course, there's no getting away from people having opinions, but then I see no need to either. What might be helpful though, is a serious widespread effort to first, start paying attention to the degree to which sub-optimal thinking may be taking place in our ~societies (global, national, regional, forums, etc), and then try to figure out of anything can be done about it.

A few examples of such sub-optimal thinking/behavior that are common (start looking for it and judge for yourself) even here on HN:

- mistaking opinions for facts (at least when it comes to conversational presentation)

- the mistaken belief that true enough generalizations of a group necessarily carries down to individual members & line items - ie: since <party/candidate x> behaves in a certain way, then therefore it logically follows that each individual who voted for (or continues to support) that entity will act in unison with each other and the greater entity, on all individual decisions. Of course, no one actually explicitly thinks this way, but start paying attention and see if you notice anyone behaving this way. Typically, this manifests as an apparent perceived ability to read minds and predict the future.

- operational (behavioral) unawareness of the difference between ~facts and perspectives on facts. Or, evaluation of a high dimensional object on a subset of dimensions, without realizing it. This will often manifest as two people arguing about a complex issue and both behaving as if their position is absolutely correct....it's because both positions very well may be correct!

- "believing" (due to heuristics) that because something worked out one way in the past, it will always work out that way in the future, regardless of whether the conditions have changed (ie: socialism/capitalism will result in <x>)

- an apparent inability/unwillingness to accept more than two states of knowledge: true or false. Not only is there a third state, unknown, but varying levels of reasoned certainty are also possible (this one seems rather innocuous, but I have a feeling it is one of the most impactful - it's difficult to argue passionately when you have no strong beliefs)

- letting emotions override logic, often in response to triggers such as inflammatory comments

- and so on, the list is endless

> Trying to find the minimal common position is probably more constructive and you would need to work out things from there. Conflicts will arise inevitably that need to be dealt with by compromises.

Exactly, but a prerequisite for that is the development of skills in understanding how the human mind works, and an ability to put that knowledge into practice. Many intelligent folks have decent academic skills in this area already, but change the topic to something controversial and watch the memory of those skills vanish as they revert to heuristic-based thinking.

> As for vetting information, there simply isn't any authority. Everyone could become one, you would just need to convince all the others.

I completely agree, but how many people actually realize this on a behavioral basis, whether or not they believe it on an academic basis? In a thread about trustworthiness of news, there will typically be little more than minor disagreements here, but move the topic to something else, that is based on the content broadcast by that very same news, and watch the fireworks start. This is the phenomenon I refer to as operational/behavioral, or putting the knowledge into practice. Like Mike Tyson says, everyone has a plan until they get punched in the mouth.

The general sentiment I perceive on whether improvements can be made in this area tends to be quite negative. It's been like this forever, so why should we expect it to ever change? Well, have we ever actually tried to do something about it? The US government and media claim there is a near crisis level of disinformation based foreign interference in our democracy, yet have you heard any politician float the idea of maybe putting a little effort into teaching people how to think, just as we teach them all sort of other things that don't happen to come naturally (math, spelling, physics, programming, etc). It seems like a rather odd situation to me.

Or, rather than sitting around waiting for the government to do something on a grand scale, is there a smaller domain or platform where individuals could collectively agree to start working on these issues, sort of as a group experiment in mass psychology?

> Key takeaway: It is impossible to have a neutral arbiter that is also perceived as neutral.

Actually, it is impossible to have a neutral arbiter. The problem isn't one of perceptions and allegation of bias. Most statements either are not objective (Trump is a bad president) or are objective in theory but cannot be determined (Trump intended to dig up dirt on his political rival). Nature doesn't provide a zero detent, so there is no way to be a neutral arbiter.

The problem with Big Tech determining what is and what is not fake news isn't censorship, it is that they have proven time and again that they are no better at determining what is true than we are. Big Tech doesn't have some rigorous technique, they pass the buck to media companies. Media companies pass stories along like Chinese whispers... nobody bothers doing investigative reporting anymore because it's too expensive given the shifts in the marketplace. So on a per-story basis, only a handful of people actually have direct knowledge, and as humans with biases they relay their interpretation of that knowledge. Everyone down the line parrots the bias unknowingly. The only way we find out the mainstream media is wrong is when another person with direct knowledge took perhaps a video and posts it on YouTube. Unfortunately for that poor sod, he gets the video taken down as fake news.

> extrapolates from the fake news crisis that started in 2016

That's fake news. A fake news crisis was popularized in 2016, around the election of a particularly controversial politician. But I've seen no evidence that it started or even got much worse then, rather than just becoming more visible. And it's worse when we're not aware of it, so this "crisis" may have actually been progress in that it deservedly decreased trust in news.

Of course there was disinformation ("fake news") before 2016 -- nobody would dispute that -- but it wasn't really framed as a crisis yet. So I would suggest you don't just dismiss this whole essay on the basis of semantic squabbling about the tl;dr summary.

Who has to frame it as a crisis for it to count? I remember being very disturbed by the fake news getting upvoted on reddit when I was in high school. I remember talking about it with other people who were also disturbed.

When you were in high school Reddit accounted for maybe 5% of the users that Facebook and Twitter have right now

Most of them were not moms or grandmothers

I remember some fake news about Iraq and WMD around 2003.

It wasn't news, that was convenient untruth.

Up until two years ago I was still able to find a NBC or CBS article from May 2003 saying how more than 80% of the US public was approving of said military intervention and how that was making the whole thing ok (or at least that was the implicit message).

I used to link to that article whenever this subject of fake news (or propaganda as it was called at the beginning of the 20th century) came into discussion on reddit or here, on HN. A couple of times I also linked to the Athenians' pre-Sicilian Expedition [1] propaganda campaign led by a certain Alcibiades, a war which had disastrous consequences on Athens itself. So it seems that we have been in an "information apocalypse" for at least ~2400 years now.

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

There are much more recent examples. Venezuela, "burned humanitarian aid trucks" fiasco.

>A fake news crisis was popularized in 2016

I think it's reasonable to characterise that as a fake news crisis. I'm fed up with pedants insisting that every mention of everything has to be annotated and caveat-ed with every possible qualification. If you want to add context or broaden the discussion, there are positive ways to do that which add value other than a whiny "well actually" post.

It's a crises because establishment media now have difficulty shaping public opinion through the traditional mechanism of editorial choice.

Technology is before culture and politics. Without the printing press , the nation state wouldnt be possible, and without the elite-controlled mass media, the centralized bureaucratic machine which monotonically increased its power concentration in the late-20th would not have been possible. The internet disrupted that, and now we have on the one hand states with a legacy of Too Much Control on people 's lives and on the other hand citizens who are disengaged and disinterested , again because of technology. Your neighborhood is dirty and full of bigots? That's OK because you live through a shiny screen anyway and you get to chat with uber-progressive friends every day. Your city looks ugly? That's OK because paradise beach weekends are only a flight away. And that's progress because it allows us to be increasingly atomized and tolerant of our different neighbors. But we 've forgotten all that legacy power which is concentrated in the hands of Leaders, and while people are distracted, bad actors are going to be taking advantage of it. Fake media and underhanded tactics are going to be rampant until we begin to remove or decentralize powers.

Fake news didnt start in 2016.

I had an entire semester on media manipulation for my sociology degree in 1995. Propaganda has played a major part in democracy since the invention of the printing press. Its good that there is an increased awareness of the need for critical thinking, but to say that we are currently living through a "fake news crisis" is wrong

It is certainly worse since everyone could publish themselves easily and quickly. At the time of paper papers, where you actually needed access to a printing press and physical distribution network, there were much fewer actors, and each of them had to take at least some care of their reputation. Not to mention that they were easy to localize, so they could actually be sued for the most blatant smears.

But then of course, propaganda and misinformation are probably as old as mankind. It's just that the scale has changed.

Could be a tech solution to the reputation problem? Some kind of distributed social media- and wikipedia-like system where information is signed by its author(s), and where people are able to produce metadata rating the factuality of information and authors.

Currently people can create such information, of course. But liars routinely deny well-documented things in their past, and those denials are effective to their intended audience. And currently someone can debunk a lie, but there's no systematic way to find everyone who testifies that it's untrue.

And while finding all of a claim's supporters and debunkers would be a great first step, a truly good (i.e. skeptical) reader would want to be able to find their arguments, too.

I'm skeptical that current AI could automate the job of indexing that information -- understanding controversial arguments seems likely to require true intelligence. I happen to have made a system[1] in which indexing any kind of data or metadata is extremely easy, requiring little more than knowing some natural language. (Deciding what kind of information you want to encode is hard, but once you've decided that, encoding it is easy.)

[1] https://github.com/JeffreyBenjaminBrown/hode

I don't think the reputation problem is particulary hard or need any fancy technical solutions. If I read something on NYTimes.com I know that the information there is true most of the time (even thought it do happen that they screw up). The job of good journalism is to vet their sources, finding collaborating evidence and listen to those that have different claims. If I read something from a source I don't know anything about, I must keep in mind that what I'm reading just might not be true.

Absolutely, that is the job of journalists.

But if we face an information apocalypse, it's because many people can't distinguish between good and bad journalism. If the data journalists base their work on was open, I believe we could do better.

> If I read something from a source I don't know anything about, I must keep in mind that what I'm reading just might not be true.

Indeed, that's what you should do, given current constraints. You shouldn't verify by repeating the work the journalist did, because you'd never have time for anything else. But you shouldn't have to repeat it -- it could be available. They already did the work; all they'd need to do is encode it in a readable, searchable way.

Moreover such a system would let us leverage far more peoples' input. While the barriers for entry have fallen sharply, it's still hard to build a readership, because it takes a while to convince people to trust you. If it were easier to evaluate the output of budding bloggers, they wouldn't have to waste a decade shouting into the void before making a difference. (The death of local news is a big problem. This might help, among other problems, that one.)

And, hopefully, through the same mechanism, it would become harder for a propoganda outlet to establish themselves.

This simply doesn’t work because “information”, in the form of news articles and blog posts, consists of 10% actual facts and 90% narrative framing. As such, any distributed rating of factuality will instantly devolve into a rating of popularity, because modern information is simply not made of facts. And we already have popularity ratings and they don’t work for truth.

> news articles and blog posts, consists of 10% actual facts and 90% narrative framing

Agreed! That's part of my motivation. A federated knowledge graph would allow people to distill the kernels of insight from the fluff and share the good stuff.

> we already have popularity ratings and they don’t work for truth.

I envision no algorithm to rate truth. I'm skeptical it's possible (but interested in suggestions). I just want people to be operationalize their definition of truth. I'd like you to be able to search for economic forecasts by people with PhDs from Harvard made prior to the last interest rate hike. Or statements regarding the West Bank agreed upon by a majority of Israelis and a majority of Palestinians.

Or incoherent rates by an orange man-child with his head up his ass. I can't guarantee a good index and search engine would be used the right way. I just think it would be better if we could get our hands dirty, rather than hoping Google does the right thing.

Part of my aim is that the data behind counterarguments be as easy to reach as the opinion they led to. It's one thing for me to see that 30% of users think vaccines cause hemmorhoid-face. It's another if I can see that their beliefs can all be traced (after factoring out cycles within that community) to a single no-longer-licensed proctologist.

The unsaid thing about fake news is that it's manipulation by non-giants. It's not the big media corps/newspapers lying to the people, it's some small website playing up cognitive bias.

That's not actually unsaid. Manipulation and falsehood by "news" sites on the web is what the "fake news" phenomenon originally referred to, but the term was successfully co-opted by the Trump campaign to refer instead to the mainstream media, and to discredit any negative reporting on him.

Those grassroots lie-factories could be more dangerous than state-directed ones. The state has a single message to push. A market, by contrast, will try every damn thing it can think of, and some of it will work.

I don't remember where I read it, but I believe the 2016 Russian disinformation campaign took advantage of a number of conspiracy theories that arose "in the wild".

I believe the influence of the Russian disinformation was overstated that it could very well be declared fake news.

Which messages? Which groups were targeted? How many people did it hit?

It was targeted to increase distrust? On which platform?

What historically was regarded as propaganda could very well be dissected afterwards. I don't really see that here or any arguments that this influence cannot be traced.

I would be very surprised if Russia didn't try to boost their favorite candidate as a geopolitical rival, but to imply relevant Russian interference is an example of distorting the truth in my opinion. And not a trivial one.

Furthermore you are also implying that everyone having a somewhat dissenting opinion about unrelated issues to being manipulated. Some might take issue with that.

Not sure I agree with that. The difference between CNN and Fox, say, is far larger than the difference between the "big three" was back in, say, the 1970s (the time of Vietnam, and the US government trying very hard to paint a world that was different from reality). What we have now is the major media trying to paint a world, rather than just report it. In any given instance, that's not necessarily "fake news". But by what you choose to cover and what you choose to ignore, you can paint a very different picture. What you say is still true, but it doesn't add up to the truth. It presents a distorted picture of the world.

Religion has been around a fair whack longer than the printing press.

I saw a very interesting talk by Mike Tamir, Berkeley and Uber Autonomous Group, about an extension his research group built to detect sensationalism and emotional appeals. The idea was that detecting writing designed to anger, incite and sensationalize (versus state facts) basically separated everything we’d call fake news from journalism: https://www.fakerfact.org/about

This would only detect the crudest type of bias —- it does absolutely nothing against selective reporting or editorial slant. People read sensational news precisely because they think “respectable” outlets are selectively and disingenuously deciding to ignore it.

I think this is an important point. I am reposting it here since it is at the bottom of the page and we don’t always get there.

> And so, maybe the majority of people wouldn’t even want SafeBrowsing-style blacklists of fake news sites or verification badges on legitimate journalist-vetted news articles, because they’re not reading the news to learn the truth - they’re reading the news to validate and spread their existing worldviews.

>Effectively this means that any technological solution to the information apocalypse depends on a social/behavioral solution: people need to welcome cognitive dissonance into their online spaces instead of shunning it. But it sounds almost ridiculous to suggest that shares/likes/retweets should be based on factual accuracy, not emotions. That’s not how social media works.

The idea of a reliability-indicator has been floating around for some time for few-to-no viable results. A much more broad-spectrum immunity-system building exercise is Crash Course's Navigating Digital Information ( https://www.youtube.com/watch?v=M5YKW6fhlss&list=PL8dPuuaLjX... ), which is excellent, and the ep linked is strongly recommended for everyone.

I have experience in an unusual domain that's shaped the way I think about truth and misinformation. I play super smash bros melee for the Nintendo gamecube competitively, and I view it as an odd sort of sanity test for the limits of truth.

The deck is impossibly stacked in favor of truth in melee. To start off with we actually have Truth with a capital T. Press Y and dpad down and you bring up white text telling you exactly what state your character is in. Imagine being a psychiatrist and instead of the messy uncertain process of diagnosing patients you can press a button that freezes time and white text from a debug menu god forgot to remove from the public release of life appears. Time switches to frame advance mode and "DepressionModerate 27" floats over your patients head.

Not only do we have the truth, but we have a community that values performance and has a vested interest in the truth.

And not only that, but there isn't even political or partisan resistance to the truth.

Ok, so this is a world where pure objective truth exists, science tools are free easy and available to all, the populace cares about truth, politics doesn't exist, there's no real incentive to spreading misinformation, and there are no large scale or individual actors purposely spreading misinformation. And yet even in this best of all possible worlds microcosm truth always seems to be barely hanging on by the skin of its teeth.

There's always someone around the corner who says that powershielding an attack incurs no shield stun.

I don't think people appreciate just how fragile the truth is even before adding in bad guys. I think we fall too quickly into viewing truth and misinformation through the most exciting narratives. Like getting in a frenzy over shark attacks and neglecting the thousands of people killed by the boring old flu.

I'm not so sure about the incoming "reality apathy". People seem to be more and more ready to mobilize, the more pervasive social media has become (completely saturated already, yes). That applies to both sides though. And to all issues - from the important to the petty.

It's not so much that they don't care about their beliefs enough to act on them; it's that they don't care about verifying those beliefs.

The solutions proposed don't understand what can and cannot be verified. You can verify if a statement is actually from the source subject to the messy details of distribution. You can't verify the actual truth cryptographically. Ans ot encounters the same messy problems as computer security - no matter how good your technology you can't have both power to the user and protect from absolute idiots - see phishing.

An authority signing it only tells you that it was approved by one and not that it is actually valid. Transparency enabled by the internet and sousveillance (for instance every camera phone capture of police brutality) makes it clear that the idea you can trust and rely upon authorities is a flat out lie.

i think the alarm is a bit overrated. it talks about "misinformation campaigns indistinguishable from reality", but if so it means they don't add more view prejudice to reality than the one that it already possess, so no harm

the danger is when disinformation distorts reality and in that case it's very distinguishable from reality, at least, i think, to an educated mind. for the others i don't think misinformation does anything for their worldview than what they have right now, if anything is just a case of confirmation bias playing on their susceptible minds

the greatest danger is for people with open minds but little literacy to digest and filter info from crap noise that can become prey of some well crafted campaigns to elude. but for better or worst people are more educated now than they were once before and we can foresee that they will be even more when the problem becomes too manifest, but will it be enough?

but even in the worst scenario i anticipate probably nothing more than an inconvenient drawback - but there's how society evolves anyway, a perpetual dialectic motion of advances and set backs - because to be really useful prejudice world views should possess better predicaments for the worlds troubles and that they have not, because paradoxically if that had they would be not prejudices in the first place. so they could only engage people on some conditioned response temporarily before reality throws them out. but maybe i'm being naive, who knows?

Bonsai kitten made me sad :-(

Also - the author uses GMail. That's an information apocalypse right there - the apocalypse of large commercial corporations and the government spying on your personal correspondence.

Any possible solution I can think of to this problem of trustworthiness relies on some notion of a trusted arbiter.

As an example: We could replace e-mail with an identity-verified communications platform. If we do that, you’re sacrificing your privacy AND placing your trust in whatever government or company manages that system.

Likewise video. A government could realistically just release statements through their own websites (e.g. Gov.uk or whitehouse.gov) and make everyone take their news from there.

In either case you have to place your trust in someone to tell you what’s true, and if you do that, the platform becomes a propaganda machine about 15 minutes after launching and we’re basically a communist state.

Basically the solution to these problems is even worse than the already terrible status quo.

This is real, and it’s here now, and Joe Public has no idea why everyone doesn’t love Trump/Bernie/Boris/Jeremy as much as he does.

This doesn’t end at all well.

I’m going home to hug my kids.

Before cryptocurrencies, we believed we needed a central authority to create money, too.

We currently have to trust arbiters of fact because we have no systematic way of evaluating the work of journalists, nor of aggregating the distibuted verification efforts that millions of people perform daily. If that work were easily navigable, our collective intelligence might improve.

In a previous comment[1] I suggest a software solution. I've already written part of it.

[1] https://news.ycombinator.com/item?id=21664084

I had a read at this but I’m not sure how reliable consensus is for something that is ultimately unprovable.

With cryptocurrency I have mathematically provable truths to rely on. With consensus I’m relying on collective opinion.

If something sounds believable to most people it would be accepted as true by any such mechanism. If the last four years have taught us anything, it’s that this is a terrifying prospect.

I'm not proposing a consensus mechanism. Rather, a mechanism that allows anybody to encode anything, and anybody else to run searches over it, systematically. It would make it easier to inspect the data behind anybody's opinions, and thereby, I hope, make us collectively smarter.

Yes, systems to encode and query data already exist, but what they can encode is limited, and the way to encode or search for it is difficult. It's hard to think of information that Hode cannot encode, and the way of getting data into or out of it is extremely close to ordinary natural language.

> We could replace e-mail with an identity-verified communications platform.

We already have that, it's what instant messengers do. Instead of accepting messages from everyone, approve each sender. Spam filters will adapt to reject the most-highly-rejected accounts. And there doesn't even need to be identity verification.

> could realistically just release statements through their own websites

Or they could verify the checksums of their videos in a public blockchain. no need for centralized media

Chris Hayes's podcast episode "The Antisocial Media" is pretty interesting, in this regard.

Direct link with downloadable mp3 (otherwise hard to find):

https://podbay.fm/podcast/1382983397/e/1569308400

Audio: https://www.podtrac.com/pts/redirect.mp3/traffic.megaphone.f...

> ... any technological solution to the information apocalypse

Here we go again. Propagating the unbroken (and unvalidated) belief that non-technological problems must have a technological "solution".

all solutions are technology

Information Highway to Hell

HN: "You can't talk politics here!"

Then I see this on the front page--which is political as hell. That any of this is a "problem" that needs to be "solved" is a limousine liberal fiction--pulled out of thin air to preserve their sanity in the face of a Trump presidency.

Michael Moore called it before the election--Trump was going to win because the Democratic party under Clinton and Obama abandoned the working class. Not enough money to address Flint, oxy, and joblessness--but somehow, there's always room on the credit card to bail out Goldman Sachs, or destabilize another middle-eastern country.

Reply on news.ycombinator.com