Comment by cm2012
1 day ago
Standard PSA: Cambridge Analytica electing Trump is not a true story. Cambridge Analytica absolutely broke Facebook's terms of service with its personality quiz scheme but it did not affect the 2016 election.
I'm an advertiser who has managed hundreds of millions of dollars for small startups and big corporations, brands, as well as a few small political campaigns.
Even at the time of the Trump election, the kind of ad targeting that Cambridge Analytica says they were offering through the personality quizzes was absolute garbage. There is a reason why the official response from Boz doesn't even mention the personality quiz stuff, because it was essentially irrelevant to the Trump campaign's performance on paid advertising on Meta. And I would go further to say that, except for his success in fundraising, all of his Meta advertising probably didn't move that many votes.
Cambridge Analytica, the organization, of course, loves to claim that they're the dark evil viziers behind the Trump campaign and they've advertised themselves as such to dictators all over the world. But that's just like Edward Bernays writing "Propaganda" back in the day, it is a way to sell their services. Trump digital director Brad Parscale said the campaign did not even use Cambridge Analytica’s controversial psychographic targeting.
Not to mention, the only reason that Cambridge Analytica had that open graph information available is because communities like Hacker News begged again and again for Facebook to open the graph and not have a walled web. After that they cut off that graph access.
Cambridge Analytica didn’t break the TOS.
If you go find the original ones it’s all about defending Facebook. Basically you can’t give our data to a competitor or build one. There’s a reason they had to quickly release a new ToS and its cause CA didn’t break it.
My understanding of the story [1] is the CA claimed it was collecting data for psychology research, then turned around and used the data for political consulting.
> Cambridge Analytica then arranged an informed consent process for research in which several hundred thousand Facebook users would agree to complete a survey for payment that was only for academic use.
1. https://en.wikipedia.org/wiki/Facebook%E2%80%93Cambridge_Ana...
I mean the data was collected in 2013 and then used in 2016 for political consulting. I have to imagine the guy actually did try doing psychology research and then realized it was a dead end (or that the psychology pointed out you could predict voters).
> the only reason that Cambridge Analytica had that open graph information available is because communities like Hacker News begged again and again for Facebook to open the graph and not have a walled web.
This isn’t true. All the early Facebook access revolved around engagement and anything even remotely related to data portability was verboten.
Anything remotely related to data portability was verboten?
Here is Facebook's own CTO in 2010 on HN itself saying “First, we genuinely support data portability: we want users to be able to use their data in other applications without restriction.”
https://news.ycombinator.com/item?id=1440154
> Anything remotely related to data portability was verboten?
Yes. Please read the entire sentence before replying:
> All the early Facebook access revolved around engagement and anything even remotely related to data portability was verboten.
2010 is years after the Facebook platform was launched.
Regardless, as people in that thread pointed out at the time, what he is referring to is not data portability. If you were building on the Facebook platform at the time, you would know that apps had pretty much free reign to soak up the personal info but the main thing people cared about in terms of data portability was the ability to use the social graph outside of Facebook. Facebook hated the idea because it would allow competitors to bootstrap their own social graph without having to find and re-add all their friends, so it was the one thing that everybody knew could get you booted off the platform.
1 reply →
Are you ok?
Facebook and Google were one of the first major online services offering an "takeout" feature in service of data portability. FB had multiple rss feeds. It's graph API was around quite early due to demand.
> Are you ok?
What do you mean?
> Facebook and Google were one of the first major online services offering an "takeout" feature in service of data portability. FB had multiple rss feeds. It's graph API was around quite early due to demand.
There was a period of time years before the Graph API launched that was all over the place. Facebook couldn't decide how they wanted their platform to work, so they cycled through a bunch of ideas like APIs, FBML, etc. before “Operation Developer Love” and the Graph API happened. Until that point, you could access a load of personal info, but the data portability everybody cared about – the social graph – was off the table because competitors could use it to bootstrap their own social graph without users having to find and re-add all their friends again. You tried to mirror that, you risked losing your account.
My understanding of the internet is Google is extremely flawed but fair and they didn't try to hold you hostage
Facebook is not fair and they actively do (ever try deleting an account? surprise! it's back a few months later. look at how many users we have, as measure of our success)
There's this for example https://news.ycombinator.com/item?id=1890164
Quizzes had like an 85% engagement rate on Facebook, they were legendarily effective paid eyeballs content. Adtech firm System1 aka OpenMail even merged with an adtech firm that owned a quiz content company, it made like $1m/2wks at small scale. I don't really know what Cambridge Analytica concretely did (to this day I have not seen a single screenshot of any of their creatives or a single concrete report of theirs) but I do know quizzes were really effective. For whatever purposes.
Yes they had a high engagement rate (due to how the platforms were structured at the time and how hyper-shareable it was) and you could make some money serving ads on quizzes because they had a high engagement rate. They were absolutely terrible as an ad targeting methodology.
Have you seen one of their quizzes?
I don't know if they were making an ad targeting solution. An ad targeting and voter targeting solution are different things.
All I am saying is I'd reserve your professional judgement until you've actually seen what they were doing.
2 replies →
You may be missing the actual crimes.
You're right that the tools were hype, and that they didn't work. So why, despite knowing that Cambridge Analytica tools were useless, did Trumo keep paying it?
Well, Robert and Rebekah Mercer were primary backers of the Trump campaign, and they also owned a major stake in Cambridge Analytica. Giving something like $49 million dollars to Trump and Co.
What are the odds they made a profit on the deal?
There were actual crimes but it was essentially the Russian government hacking email campaigns related to Hillary Clinton and drip-feeding them to the press. On the ad side there's none that I'm aware of.
From the article: "Jurors also found Facebook misled the public about investigations into data brokers following the Cambridge Analytica scandal, finding the company liable for over 2 million violations."
A jury just found that there were more than 43 million crimes (total) committed on the ad side. I think you're saying that the illegal data collection wasn't related to the advertising end, but it was. That's why they did it. Whether it was actually useful or not is irrelevant.
More importantly to me, the billionaires using it to launder their bribes to the US president probably won't ever be convicted, but that does not mean they aren't just low class crooks who can afford good lawyers.
[dead]
> it did not affect the 2016 election.
This is false.
> all of his Meta advertising
You don't appear to have any understanding of how Cambridge Analytica affected the election.
https://en.wikipedia.org/wiki/Cambridge_Analytica
That’s just a link to full Wikipedia page for the company. I must admit that I would be personally interested in seeing a vaguely scientific analysis of the effect they had. For all the talk on the subject I have yet to see any real analysis backed by data on the subject.
I know what it is. One can read it and see that CA's effect wasn't just "all of [Trump's] Meta advertising".
1 reply →
Nothing in that Wikipedia link disagrees with me at all.
This guy made the exact same comment 12 days ago ... that one was flagged and is dead.
Yes and I thought it was very unfair that I was flagged for taking a lot of time and effort to make an accurate comment.
1 reply →
[dead]
> Cambridge Analytica, the organization, of course, loves to claim that they're the dark evil viziers behind the Trump campaign and they've advertised themselves as such to dictators all over the world.
Aren't they defunct [1]?
[1] https://en.wikipedia.org/wiki/Cambridge_Analytica#:~:text=De...
Yes but the founders and related people are still very happy not to disabuse the notion of the Cambridge Analytica scandal.