> I don't want Google to be the judge of the validity of claims as it's perfectly possible to make an add which is 100% made up of technically correct statements but presents a misleading narrative.
What do you find objectionable about a low-pass filter that simply removes obvious lies?
The fact that I don't trust Google (or any other large corporation, for that matter) to judge what is and is not an "obvious lie" given the right political incentives. In this case, outright banning would be better and remove any ambiguity or opportunity for shenanigans.
Not that an outright ban is a good solution either, as there is a large discussion to have about what is and isn't political. Some people see some topics as inherently political, while others would say that they have nothing to do with politics.
So deciding what is and isn't banned becomes an inherently political statement.
This makes the definition of "politics" meaningless. Besides, it's a lot easier to nail down an objective definition (at least for Google's purposes) of what falls under a specific subject (ads are sold on keywords) than it is to nail down an objective definition of "truth".
You can exclude ads that mention or are paid for by any PAC, candidate for public office, or are advocating for or against any particular public policy decision, or mention the topic of elections or voting.
This would eliminate the majority of them, leaving bad actors that don't care about Google's requirements anyways. Another option would be to simply make certain keywords unpurchaseable.
Recognizing the truth or falsehood of a statement or series of statements can be complicated, but we're doomed as a species if an objective definition of truth stops us in our tracks.
When it comes to things in Google's wheelhouse, at least. Their entire advertising platform is based on categorizing ads and where they're supposed to display. This is an entirely different ball of wax than trying to determine if the text of an ad is true or false, and that's before we get into the "true but misleading in context" quagmire.
Besides, we have precedent here. Have a look at their policies surrounding what's allowed in health ads[1], they're pretty extensive, objective, and granular.
You shouldn't trust any large _institution_ to prevent "obvious lies" from spreading. Institutions are just collections of people, and there's barely a single one out there that's managed to completdly avoid the fact that people are, generally speaking, dishonest and dumb.
You'd trust The Economist (or the CBO) to be the arbiter of what other people are allowed to say? That's what my comment says, not "trust any institution to avoid telling obvious lies" (that being said, I've read some pretty dishonest stuff in the Economist, despite generally being a fan).
>You'd trust The Economist (or the CBO) to be the arbiter of what other people are allowed to say?
They already do that with their own employees. They censor and control what statements they print to control for accuracy. Journalism concerns in part determining what true statements we should print. I think the Economist and CBO do a good job of separating the facts from the fiction. I would trust them to fact-check statements. I would not trust CNN or Facebook or the Democrats to fact-check statements.
This is a good point that runs a little deeper than some people appreciate. (moving a topic into the category of "political" can already, today, be a subtle form of censorship in the public sphere)
At the same time, if we're talking about advertising, in the US we can just rely on what constitutes political advertising under the law.
I agree that would be an easy move for Google to make.
I would imagine that there would then be an incentive for potential donors to, rather then giving cash, pay for messages that fly under the radar but are designed to support a particular campaign.
I guess that just depends on the axis orientation...
But there is nothing inherently wrong about removing pure garbage ads.
Of course, one might object that it would give more visibility to harder-to-discern bad ads, and make the consumer less wary of ads if they are less used to see them with obviously wrong information. Double edged sword as always, and I would certainly ban them. Or every ad, for that matter. But that's google, it is their business model, so what are you expecting?
If China says the majority of Hong Kong support immediate reunification, backed by a statement from the Hong Kong government, is that an "obvious lie"? It comes from what is considered to be an authoritative source of truth.
If Google can overrule a government's statement, by disagreeing with China, does that mean they can overrule the official statement of any government?
That gives them the power, and incentive, to approve some statements that come from the White House, and not others that might damage their own reputation.
Or rather, knowing Google, approval of these goes through a machine, with unknown bias, and suddenly Google is only approving right-wing messages. All Democrat messages get disapproved. Or vice versa.
It places Google in position as the ultimate source of truth, without giving them the responsibility that comes with that.
On the other hand, if google replies to the query "labour manifesto" and the top link is an ad to "labourmanifesto.co.uk" that is actually run by the conservative party, that's the same category of fraud as if I entered "chase bank" and was sent to a phishing site. Allowing that ad in response to that query is defrauding voters.
Being the largest source of information in the world while being completely indifferent as to whether that information is true or not is an amazingly nihilistic, amoral position to take.
I think the most sensible thing for any ad agency to do is to ban _all_ political ads.
It may become difficult for them to differentiate what is and isn't a political ad (an ad for a gay dating app could be construed as a political ad in some nations), but they need to remove themselves from any seat of responsibility.
Based on your questions and comments I get the feeling I am communicating with someone who does not know that journalism exists or how it is supposed to work. What constitutes an "obvious lie"? is not even close to being an intractable problem. Reporting what China said is not the same thing as putting it forward as the truth. Determining whether it is the truth or not is not necessarily an impossible task. Etc. etc.
It's true enough that Google will experience some pressure to kowtow to the governments or powerful interests who have the ability to lean on them, and that is probably a pretty good argument to get out of the business of political advertising altogether. "What is truth anyway, man????" is not much of an argument.
> Based on your questions and comments I get the feeling I am communicating with someone who does not know that journalism exists or how it is supposed to work.
I'll try not to take offense at that.
I'm just someone who has lived under an authoritarian regime where what is "truth" has been actively manipulated by the powers that be - dissenting opinions were made illegal.
> "What is truth anyway, man????" is not much of an argument.
That wasn't the "crux" of the argument.
No advertising company should be in the business of deciding what to report as truth. If they are, complying with local laws about what is "truth" may in fact be completely antithetical to the pretense of fact checking.
The most sensible solution for someone like Google is banning political advertising altogether - not putting themselves in a position where they may be required to act as propaganda mouthpieces.
What do you find objectionable about a low-pass filter that simply removes obvious lies?