[BLKNSLVR]

I don't understand your position at all (and I don't mean this aggressively, I just can't get myself into a position of understanding). From my current perspective you seem to be absolving Google of any responsibility for, literally, their product.

My reasoning is:

Google have created an advertising platform. It is their raison d'etre; their entire millions per hour profit engine. But they only built the easy, profitable half because there were / are no regulations to enforce the responsible, difficult half.

They built the half that allows anyone with the money to put something on their platform. They didn't build the half that makes sure that they're not helping scammers, con artists, and outright criminals from reaching the global audience that their wonderful, profitable, scalable platform enables.

They should be policing it on their own to an extent that obvious scams and fake banking websites and clickbait should be detected. Even just to appear to not be a crime-facilitation platform, which they currently are.

To me it feels analogous to Microsoft and their commitment to security of Windows. It's not a priority because it's counter to profitability. Privatise the profits and socialise the costs.

If they can't control their own platform, they should not have the platform. It is not a mature enough product to be released upon the world. It is Frankenstein's monster, left to roam.

I would be supportive of legislation that outright banned advertising with Google until they were able to provably clean up their act; if their product was market-ready.

Which three letter agency said that an ad blocker was a required layer of security when browsing the web? There's a good reason: Google. If the Internet is full of scams, who is most responsible for its proliferation?

(I have a massive bias against advertising, so that heavily colours my opinion. I also understand that advertising is inevitable, but it should be held to a much higher standard than, well, the none that exists)

[WorldMaker]

Newspapers and TV channels were held to Truth in Advertising Laws, and obvious scams would trigger law suits, from consumer groups, from the FCC itself sometimes, from rival newspapers and TV channels.

Google has enough of a near-monopoly in ads that there aren't rivals scaled enough to defend the public good out of the greed of their pocket books if not the good of everyone else. Google's competitors are too busy selling ad space to the same and similar lowest common denominator polluters and scammers than to police Google enough to hope to put it out of business.

The FCC gave up on policing anything on the internet ages ago. Between Congressional protections and multiple administrations trying to keep the FCC weak, it doesn't seem to have enough power to do anything towards Google.

I suppose that leaves consumer advocacy groups. I don't know how we build and scale a Consumers Against Predatory Google Ads to the point that it can force Google to reconsider their approach to advertising for the consumer good instead of evil. But then I also don't understand how Google isn't already doing tremendous enough damage to their brand that there isn't already enough of a sentiment that Google is actively engaged in evil and needs to stop, five or six years ago at least (or maybe at least as far back as when Google bought/merged Doubleclick). Their brand, too, seems too big to fail at this point, and it is weird to me.

[AnthonyMouse]

> Google have created an advertising platform. It is their raison d'etre; their entire millions per hour profit engine. But they only built the easy, profitable half because there were / are no regulations to enforce the responsible, difficult half.

Baked into this is the assumption that the "responsible, difficult half" is something that they, rather than law enforcement, should be doing to begin with. Which they can't do effectively because they don't have the ability to impose significant penalties and we certainly don't want to give them that.

Consider how this is even supposed to work. There is often no conclusive way to tell ex ante whether some website is a scam; the premise of them is to look like a real site. Google doesn't know if it's a scammer or some third party contractor who has be commissioned by the actual entity to set up a new site, and real sites are often full of bad UI choices and weird bugs, or open up to nothing more than a login page which the ad network has no credentials to get past.

The way this works with actual law enforcement is that someone suspects a site of being a scam, or a victim files a police report, and then officers investigate. They spend significant resources to figure out if it's actually a scam or just e.g. a competitor trying to have their competition removed from visibility. They have special powers to conduct searches with probable cause. And then if it is a scam, they arrest the scammers, and if it's a false report, they charge that person for filing a false police report, which are necessary in order to provide a deterrent and prevent both of those misbehaviors from proliferating.

Otherwise you get untold numbers of false reports from people trying to grief their rivals, it takes just as much work to do a thorough investigation for each one, and then even if you catch the actual scammers, they just immediately reappear under a different name.

So then they get zillions of reports, many of them fraudulent, and you're proposing to put them in jail if they ever fail to do something they there is no apparent way for them to consistently do.

> I would be supportive of legislation that outright banned advertising with Google until they were able to provably clean up their act; if their product was market-ready.

Consider how this would apply to anything other than advertising. Some bank robbers buy Halloween masks to hide their faces. Drug dealers sell drugs over the internet and a normal mail/package carrier delivers the packages. A hitman rides the bus on the way to a hit. These are felonies! But we don't expect UPS to cut open everyone's packages or the bus driver to strip search every passenger to check for a vial of polonium, and punish them if they ever fail, because that isn't their job. They're not the police.

[BLKNSLVR]

I understand your point a lot better now, thank you. And I like the 'public cars / roads' example and have possibly used it myself in discussions about cryptography and government backdoors etc.

Respectfully, however, I still disagree in this case, although I'm struggling how to articulate why - beyond admitting it could just be my bias against advertising.

My best attempt at articulation is:

Google have built a private road over the top of the public road, that is then spilling sewage and toxic materials onto the public road. Google's private road purely serves to profit Google. There's no public good it is serving, not like a public road or electricity infrastructure.

In writing that, yes, it's a government policy / legislation / regulation failure. But also, Google should be doing better. However, as WorldMaker said above, their brand seems somewhat confusingly, unaffected, so Google has incentive to do better than bottom-of-the-barrel.

As such I can't lay no blame, or anything less than majority blame, at Google's feet. Yes, government's around the world should be doing better at containing the spilling of toxic waste into public spaces that Google both facilitates and turns a blind eye to. Google are too profitable to fail since they can buy the continued regulation vacuum.

I'm not sure how far that went off the rails, I may have just continued ranting a bit there.

Thank you for your replies, though, I do see it from a slightly different angle.

[K0balt]

Google is not a public road. It is A company that operates on public roads that delivers advertising. It’s a company driving around billboard trailers on public roads. Are you saying that a billboard company should not have any ethical standards or liability regarding the companies that they advertise on their billboards? Somehow, Google manages to not openly advertise CSAM buttthey can’t shutdown obvious scams? It’s simply a matter of having a responsible review procedure before publishing an ad. In the age of LLMs , you can’t tell me that is impossible.

[AnthonyMouse]

> Google is not a public road. It is A company that operates on public roads that delivers advertising.

So is UPS or FedEx or USPS. Are you saying they should be responsible for the contents of every package? How are they even supposed to know what the contents is? They can ask the sender but the sender can just lie.

> Somehow, Google manages to not openly advertise CSAM buttthey can’t shutdown obvious scams?

CSAM is a lot more obvious than most scams, because CSAM looks like CSAM whereas a fake login form just looks like a any other login form.

> In the age of LLMs , you can’t tell me that is impossible.

LLMs are exactly the thing that doesn't work for this.

[K0balt]

UPS and FedEx don’t know the package content because it’s sealed and private.

An advertisement is the exact opposite of private, and Google absolutely can review every single ad that runs on their system.

I’m talking about ads, things they take money in order to make public, not encrypted data they move from point a to point b.

[AnthonyMouse]

> An advertisement is the exact opposite of private

The ad itself doesn't tell you anything. The ad copy can look legitimate. Even for many real websites, the link will commonly go through a third party redirect for tracking purposes, but then the ad network has no way to know if the third party server changes where the link redirects after they reviewed it. They also have no way to know if the site the link points to is affiliated with the entity it claims to be or not.

> Google absolutely can review every single ad that runs on their system.

You can't spend $1000 investigating an ad someone is paying you $10 to run, and spending a small fraction of $10 to review an ad is not going to be thorough enough catch most of the offenders.

[K0balt]

I think there was a a line somewhere between “we will literally run ads that purport to sell products that obviously cannot exist based on any known physics” and “we audit the financials of the parent companies of anyone that advertises on our platform” that would reasonably qualify as due diligence.

The problem isn’t that sometimes Google screws up and lets a bad actor advertise on its platform, it’s that Google (and others) have made it their business model to allow bad actors to advertise on their platform. They knowingly profit from fraud.

Out of curiosity I have followed up on some of the most blatantly ridiculous ads. They typically are a first tier filter for gullibility, and never actually lead to the promised product, but rather diverting into a web of identity theft fishing, investing fraud, cryptocurrency theft, malware droppers, etc. These are funnels into real harm. And the sucker pitch is intentionally an obviously impossible product that any decent 7B LLM would tell you is probably fraudulent.

Now that I followed a few up to see where they went, that account gets shown hundreds of ads of this type from Google, and especially on YouTube. I did some AdWords research to see what those click throughs would sell for and they are very high priced exposures and clicks.

This would be trivial for Google to flag if they wanted to, but it’s a significant and growing revenue generator.

[K0balt]

Does this appear to be an official government website?

Does this url match the government website in question?

9/10 my local 7B LLM answers correctly.

[AnthonyMouse]

A 10% false positive rate is worthless.

[K0balt]

No, it’s not. Appeal->higher level review. 97 percent of your regular customers get served with no issues. The problem is that the fraudulent ads fetch much higher than average CPV and CPC.