How Fake and False News Distort Google and Others

Views: 1258

With all of the current discussions regarding the false and fake news glut on the Internet — often racist in nature, some purely domestic in origin, some now believed to be instigated by Putin’s Russia — it’s obvious that the status quo for dealing with such materials is increasingly untenable.

But what to do about all this?

As I have previously discussed, my general view is that more information — not less — is the best solution to these distortions that may have easily turned the 2016 election on its head.

Labeling, tagging, and downranking of clearly false or fake posts is an approach that can help to reduce the tendency for outright lies to be treated equivalently with truth in social media and search engines. These techniques also avoid invoking the actual removal of lying items themselves and the “censorship” issues that then may come into play (though private firms quite appropriately are indeed free to determine what materials they wish to permit and host — the First Amendment only applies to governmental restraints on speech in the USA).

How effective might such labeling be? Think about the labeling of “fake news” in the same sort of vein as the health warnings on cigarette packs. We haven’t banned cigarettes. Some people ignore the health warnings, and many people still smoke in the USA. But the number of people smoking has dropped dramatically, and studies show that those health warnings have played a major role in that decrease.

Labeling fake and false news to indicate that status — and there’s a vast array of such materials where no reasonable arguments that they are not untrue can reasonably exist — could have a dramatic positive impact. Controversial? Yep. Difficult? Sure. But I believe that this can be approached gradually, starting with top trending stories and top search results.

A cure-all? No, just as cigarette health warnings haven’t been cure-alls. But many lives have still been saved. And the same applies to dealing with fake news and similar lies masquerading as truthful posts.

Naysayers suggest that it’s impossible to determine what’s true or isn’t true on the Internet, so any attempts to designate anything that’s posted as really true or false must fail. This is nonsense. And while I’ve previously noted some examples (Man landing on the moon, Obama born in Hawaii) it’s not hard to find all manner of politically-motivated lies that are also easy to ferret out as well.

For example, if you currently do a Google search (at least in the USA) for:

southern poverty law center

You will likely find an item on the first page of results (even before some of the SPLC’s own links) from online Alt-Right racist rag Breitbart — whose traditional overlord Steve Bannon has now been given a senior role in the upcoming Trump administration.

The link says:

FBI Dumps Southern Poverty Law Center as Hate Crimes Resource

Actually, this is a false story, dating back to 2014. It’s an item that was also picked up from Breitbart and republished by an array of other racist sites who hate the good work of the SPLC fighting both racism and hate speech.

Now, look elsewhere on that page of Google search results — then on the next few pages. No mention of the fact that the original story is false, that even the FBI itself issued a statement noting that they were still working with the SPLC on an unchanged basis.

Instead of anything to indicate that the original link is promoting a false story, what you’ll mostly find on succeeding pages is more anti-SPLC right-wing propaganda.

This situation isn’t strictly Google’s fault. I don’t know the innards of Google’s search ranking algorithms, but I think it’s a fair bet that “truth” is not a major signal in and of itself. More likely there’s an implicit assumption — which no longer appears to necessarily hold true — that truthful items will tend to rise to the top of search results via other signals that form inputs to the ranking mechanisms.

In this case, we know with absolute certainly that the original story on page one of those results is a continuing lie, and the FBI has confirmed this (in fact, anyone can look at the appropriate FBI pages themselves and categorically confirm this fact as well).

Truth matters. There is no equivalency between truth and lies, or otherwise false or faked information.

In my view, Google should be dedicated to the promulgation of widely accepted truths whenever possible. (Ironic side note: The horrible EU “Right To Be Forgotten” — RTBF — that has been imposed on Google, is itself specifically dedicated to actually hiding truths!)

As I’ve suggested, the promotion of truth over lies could be accomplished both by downranking of clearly false items, and/or by labeling such items as (for example) “DEEMED FALSE” — perhaps along with a link to a page that provides specific evidence supporting that label (in the SPLC example under discussion, the relevant page of the FBI site would be an obvious link candidate).

None of this is simple. The limitations, dynamics, logistics, and all other aspects of moving toward promoting truth over lies in social media and search results will be an enormous ongoing effort — but a critically crucial one.

The fake news, filter bubbles, echo chambers, and hate speech issues that are now drowning the Internet are of such a degree that we need to call a major summit of social media and search firms, experts, and other concerned parties on a multidisciplinary basis to begin hammering out practical industry-wide solutions. Associated working groups should be established forthwith.

If we don’t act soon, we will be utterly inundated by the false “realities” that are being created by evil players in our Internet ecosystems, who have become adept at leveraging our technology against us — and against truth.

There is definitely no time to waste.

–Lauren–
I have consulted to Google, but I am not currently doing so — my opinions expressed here are mine alone.
– – –
The correct term is “Internet” NOT “internet” — please don’t fall into the trap of using the latter. It’s just plain wrong!

Blocked by Lauren (“The Motion Picture”)

Views: 615

With nearly 400K Google+ followers, I’ve needed to block “a few” over the years to keep order in the comment sections of my threads. I’m frequently asked for that list — which of course is composed entirely of public G+ profile information. But as far as I know there is no practical way to export this data in textual form. However, when in doubt, make a video! By the way, I do consider unblocking requests, and frequently unblock previously blocked profiles as a result, depending on specific circumstances. Happy Thanksgiving!

https://www.youtube.com/watch?v=GX79fYTSjFE

–Lauren–
I have consulted to Google, but I am not currently doing so — my opinions expressed here are mine alone.
– – –
The correct term is “Internet” NOT “internet” — please don’t fall into the trap of using the latter. It’s just plain wrong!

Facebook, Google, Twitter, and Others: Start Taking Hate Speech Seriously!

Views: 1934

Recently, in Crushing the Internet Liars, I discussed issues relating to the proliferation of “fake news” on the Internet (via social media, search, and other means) and the relationship of personalization-based “filter bubbles” and “echo chambers” — among other effects.

A tightly related set of concerns, also rising to prominence during and after the 2016 election, are the even broader concepts of Internet-based hate speech and harassment. The emboldening of truly vile “Alt-Right” and other racist, antisemitic white supremacist groups and users in the wake of Trump’s election has greatly exacerbated these continuing offenses to ethics and decency (and in some cases, represent actual violations of law).

Lately, Twitter has been taking the brunt of public criticism regarding harassment and hate speech — and their newly announced measures to supposedly combat these problems seem to mostly be potentially counterproductive “ostrich head in the sand” tools that would permit offending tweets to continue largely unabated.

But all social media suffers from these problems to one degree or another, and I feel it is fair to say that no major social media firm really takes hate speech and harassment seriously — or at least as seriously as ethical firms must.

To be sure, all significant social media companies provide mechanisms for reporting abusive posts. Some systems pair these with algorithms that attempt to ferret out the worst offenders proactively (though hate users seem to quickly adapt to bypass these as rapidly as the algorithms evolve).

Yet one of the most frequent questions I receive regarding social media is “How do I report an abusive posting?” Another is “I reported that horrible posting days ago, but it’s still there, why?”

The answer to the first question is fairly apparent to most observers — most social media firms are not particularly interested in making their abuse reporting tools clear, obvious, and plainly visible to both technical and nontechnical users of all ages. Often you must know how to access posting submenus to even reach the reporting tools.

For example, if you don’t know what those three little vertical dots mean, or you don’t know to even mouse over a posting to make those dots appear — well, you’re out of luck (this is a subset of a broader range of user interface problems that I won’t delve into here today).

The second question — why aren’t obviously offending postings always removed when reported — really needs a more complex answer. But to put it simply, the large firms have significant problems dealing with abusive postings at the enormous scales of their overall systems, and the resources that they have been willing to put into the reporting and in some cases related human review mechanisms have been relatively limited — they’re just not profit center items.

They’re also worried about false abuse reports of course — either purposeful or accidental — and one excuse used for “hiding” the abuse reporting tools may be to try reduce those types of reports from users.

All that having been said, it’s clear that the status quo when it comes to dealing with hate speech or harassing speech on social media is no longer tenable.

And before anyone has a chance to say, “Lauren, you’re supposed to be a free speech advocate. How can you say this?”

Well, it’s true — I’m a big supporter of the First Amendment and its clauses regarding free speech.

But what is frequently misunderstood, is that this only applies to governmental actions against free speech — not to actions by individuals, private firms, or other organizations who are not governmental entities.

This is one reason why I’m so opposed to the EU’s horrific “Right To Be Forgotten” (RTBF) — it’s governments directly censoring the speech of third parties. It’s very wrong.

Private firms though most certainly do have the right to determine what sorts of speech they choose to tolerate or support on their platforms. That includes newspapers, magazines, conventional television networks, and social media firms, to name but a few.

And I assert that it isn’t just the right of these firms to stamp out hate speech and harassment on their platforms, but their ethical responsibility to do so as well.

Of course, if the Alt-Right or other hate groups (and certainly the right-wing wackos aren’t the only offenders) want to establish their own social media sites for that subset of hate speech that is not actually illegal — e.g. the “Trumpogram” service — they are free to do so. But that doesn’t mean that the Facebooks, Googles, and Twitters of the world need to permit these groups’ filth on their systems.

Abusive postings in terms of hate speech and harassing speech certainly predate the 2016 election cycle, but the election and its aftermath demonstrate that the major social media firms need to start taking this problem much more seriously — right now. And this means going far beyond rhetoric or public relations efforts. It means the implementation of serious tools and systems that will have real and dramatic impacts on helping to stamp out the postings of the hate and other abuse mongers in our midst today.

–Lauren–
I have consulted to Google, but I am not currently doing so — my opinions expressed here are mine alone.
– – –
The correct term is “Internet” NOT “internet” — please don’t fall into the trap of using the latter. It’s just plain wrong!

Unacceptable: How Google Undermines User Trust by Blocking Users from Their Own Data

Views: 1087

UPDATE (November 18, 2016): After much public outcry, Google has now reversed the specific Pixel-related Google account bans noted in this post. Unfortunately, the overall Whose data is it? problem discussed in this post persists, and it’s long since time for Google to appropriately address this issue, which continues to undermine public user trust in a fine company.

– – –

There are times when Google is in the right. There are times when Google is in the wrong. By far, they’re usually on the angels’ side of most issues. But there’s one area where they’ve had consistent problems dating back for years: Cutting off users from those users’ own data when there’s a dispute regarding Google Account status.

A new example of this recurring problem — an issue about which I’ve heard from large numbers of Google users over time — has just surfaced. In this case, it involves the reselling of Google Pixel phones in a manner that apparently violates the Google Terms of Service, with the result that a couple of hundred users have reportedly been locked out of their Google accounts and all of their data, at least for now. 

This means that they’re cut off from everything they’ve entrusted to Google — mail, documents, photos, the works.

Here and now, I’m not going to delve into the specifics of this case — I don’t know enough of the details yet. The entire area of Google accounts suspension, closure, recovery, and so on is complex to say the least. Most times (but not always) users are indeed at fault — one way or another — when these kinds of events are triggered. And the difficulty of successfully appealing a Google account suspension or closure has become rather legendary.

Even recovering a Google account due to the loss of a password can be difficult if you haven’t taken proactive steps to aid in that process ahead of time —steps that I’ve previously discussed in detail.

But the problem of what happens to users’ data when they can’t access their accounts — for whatever reasons — is something that I’ve personally been arguing with Google about literally for years, without making much headway at all.

Google has excellent mechanisms for users to download their data while they still have account access. Google even has systems for you to specify someone else who would have access to your account in case of emergency (such as illness or accident), and policies for dealing with access to accounts in case of the death of an account holder.

The reality though, is that users have been taught to trust Google with ever more data that is critical to their lives, and most people don’t usually think about downloading that data proactively.

So when something goes wrong with their account, and they lose access to all of that data, it’s like getting hit with a ton of bricks.

Again, this is not to say that users aren’t often — in fact usually — in the wrong (at least in some respect) when it comes to account problems. 

But unless there is a serious — and I mean serious (like child porn, for example) criminal violation of law — my view is that in most circumstances users should have some means to download their data from their account even if it has been suspended or closed for good cause. 

If they can’t use Google services again afterwards, them’s the breaks. But it’s still their own data we’re talking about, not Google’s.

Google has been incredibly resistant to altering this aspect of their approach to user account problems. I am not ignorant of their general reasoning in this category of cases — but I strongly believe that they are wrong with their essentially one size fits all “death penalty” regime in this context.

Nobody is arguing that there aren’t some specific situations where blocking a violating user (or a user accused of violations) from accessing their data on Google services is indeed justified. But Google doesn’t seem to have any provisions for anything less than total data cutoff when there’s a user account access issue, even when significant relevant legal concerns are not involved.

This continuing attitude by Google does not engender user trust in Google’s stewardship of user data, even though most users will never run afoul of this problem.

These kinds of actions by Google provide ammunition to the Google Haters and are self-damaging to a great firm and the reputation of Googlers everywhere, some of whom have related to me their embarrassment at trying to explain such stories to their own friends and families.

Google must do better when it come to this category of user account issues. And I’ll keep arguing this point until I’m blue in the face and my typing fingertips are bruised. C’mon Google, please give me a break!

–Lauren–
I have consulted to Google, but I am not currently doing so — my opinions expressed here are mine alone.
– – –
The correct term is “Internet” NOT “internet” — please don’t fall into the trap of using the latter. It’s just plain wrong!

Crushing the Internet Liars

Views: 7383

Frankly, this isn’t the post that I had originally intended. I had a nearly completed blog draft spinning away happily on a disk, a draft that presented a rather sedate, scholarly, and a bit introspective discussion of how Internet-based communications evolved to reach the crisis point we now see regarding misinformation, filter bubbles, and so-called echo chambers in search and social media.

I just trashed that draft. Bye!

Numerous times over the years I’ve tried the scholarly approach in various postings regarding the double-edged sword of Internet personalization systems — capable of bringing both significant benefits to users but also carrying significant and growing risks.

Well, given where we stand today after the 2016 presidential election, it appears that I might have just as well been doing almost anything else rather than bothering to write that stuff. Toenail trimming would have likely been a more fruitful use of my time.

So now — today — we must deal with this situation while various parties are hell-bent toward turning the Internet into a massive, lying propaganda machine to subvert not only the democratic process, but our very sense of reality itself.

Much of this can be blamed on the concept of “false equivalency” — which runs rampant on cable news, mainstream Internet news sites, and throughout social media such as Facebook (which is taking the brunt of criticism now — and rightly so), plus on other social media ecosystems.

Fundamentally, this idea holds that even if there is widespread agreement that a particular concept is fact, you are somehow required to give “equal time” to wacko opposing views.

This is why you see so much garbage prominently surfaced from nutcases like Alex Jones — who believes the U.S. government blew up the World Trade Center buildings — or Donald Trump and his insane birther attacks on Obama, that Trump used to jump-start his presidential campaign. It doesn’t take more than half a brain to know that such statements are hogwash.

To be sure, it’s difficult to really know whether such perpetually lying creatures actually believe what they’re saying — or are simply saying outrageous things as leverage for publicity. In the final analysis though, it doesn’t much matter what their motives really are, since the damage done publicly is largely equivalent either way.

The same can be said for the wide variety of fake postings and fake news sites that increasingly pollute the Net. Do they believe what they say, or are they simply churning out lies on a twisted “the end justifies the means” basis? Or are they just sick individuals promoting hate (often racism and antisemitism) and chaos? No doubt all of the above apply somewhere across the rapidly growing range of offenders, some of whom are domestic in nature, and some who are obviously operating under the orders of foreign leaders such as Russia’s Putin.

Facebook, Twitter, and other social media posts are continually promulgating outright lies about individuals or situations. Via social media personalization and associated posting “surfacing” systems, these lies can reach enormous audiences in a matter of minutes, and even push such completely false materials to the top of otherwise legitimate search engine results.

And once that damage is done, it’s almost impossible to repair. You can virtually never get as many people to see follow-ups that expose the lying posts as who saw the original lies themselves.

Facebook’s Mark Zuckerberg is publicly denying that Facebook has a significant role in the promotion of lies. He denies that Facebook’s algorithms for controlling which postings users see creates echo chambers where users only see what they already believe, causing lies and distortions to spread ever more widely without truth having a chance to invade those chambers. But Facebook’s own research tells a very different story, because Facebook insists that exactly those kinds of controlling effects occur to the benefit of Facebook’s advertisers.

Yet this certainly isn’t just a Facebook problem. It covers the gamut of social media and search.

And the status quo can no longer be tolerated.

So where do we go from here?

Censorship is not a solution, of course. Even the looniest of lies, assuming the associated postings are not actually violating laws, should not be banned from visibility.

But there is a world of difference between a lying post existing, vis-a-vis the actual widespread promotion of those lies by search and social media.

That is, simply because a lying post is being viewed by many users, there’s no excuse for firms’ algorithms to promote such a post to a featured or other highly visible status, creating a false equivalency of legitimacy by virtue of such lies being presented in close proximity to actual facts.

This problem becomes particularly insidious when combined with personalization filter bubbles, because the true facts are prevented from penetrating users’ hermetically sealed social media worlds that have filled with false postings.

And it gets worse. Mainstream media in a 24/7 news cycle is hungry for news, and all too often now, the lies that germinate in those filter bubbles are picked up by conventional media and mainstream news sites as if they were actual true facts. And given the twin realities of reduced budgets and beating other venues to the punch, such lies frequently are broadcast by such sites without any significant prior fact checking at all.

So little by little, our sense of what is actually real — the differences between truth and lies — becomes distorted and diluted.

Again, censorship is not the answer.

My view is that more information — not less information — is the path toward reducing the severity of these problems.

Outright lies must not continue to be given the same untarnished prominence as truth in search results and in widely seen social media postings.

There are multiple ways to achieve this result.

Lying sites in search results can be visibly and prominently tagged as such in those results, be downranked, or both. Similar principles can apply to widely shared social media posts that currently are featured and promoted by social media sites primarily by virtue of the number of persons already viewing them. Because — lets face it — people love to view trash. Lots of users viewing and sharing a post does not make it any less of a lie.

As always, the devil is in the details.

This will be an enormously complex undertaking, involving technology, policy, public relations, and the law. I won’t even begin to delve into the details of all this here, but I believe that with sufficient effort — effort that we must now put forth — this is a doable concept.

Already, whenever such concepts are brought up, you quickly tend to hear the refrain: “Who are you to say what’s a fact and what’s a lie?”

To which I reply: “To hell with false equivalences!”

Man landed on the moon. Obama was born in Hawaii. Terrorists destroyed the World Trade Center with jet aircraft. Hillary Clinton never said she was going to abolish the Second Amendment. Donald Trump did say that he supported the war in Iraq. Denzel Washington did not say that he supported Trump. On and on and on.

There are a virtually endless array of stated facts that reasonable people will agree are correct. And if the nutcases want to promote their own twisted views on these subjects that’s also fine — but those postings should be clearly labeled for what they are — not featured and promoted. As the saying goes, they’re free to have their own opinions, but not their own facts.

Obviously, this leaves an enormous range of genuinely disputed issues where the facts are not necessarily clear, often where only opinion and/or philosophy really apply. That’s fine too. They’re out of scope for these discussions or efforts.

But the outright Internet liars must be crushed. They shouldn’t be censored, but they must no longer be permitted to distort and contaminate reality by being treated on an equal footing with truth by major search and social media firms.

We built the Internet juggernaut. Now it’s our job to fix it where it’s broken.

–Lauren–
I have consulted to Google, but I am not currently doing so — my opinions expressed here are mine alone.
– – –
The correct term is “Internet” NOT “internet” — please don’t fall into the trap of using the latter. It’s just plain wrong!