5 ms·

Michael Lewis (Moneyball, Big Short) interviewed Alex Kogan: https://atrpodcast.com/episodes/the-alex-kogan-experience-s1!d20f3 https://atrpodcast.com/episodes

Michael Lewis (Moneyball, Big Short) interviewed Alex Kogan: https://atrpodcast.com/episodes/the-alex-kogan-experience-s1!d20f3 https://atrpodcast.com/episodes/the-alex-kogan-experience-s1... After listening to this, it sounds like Cambridge Analytica was overblown, in the sense that the information wasn’t as useful as claimed.

sonnyblarney 8y ago
It was also overblown to the extent that there as a material 'scandal'. Most of the issue really had to do with the very public fact of Facebook's API's at the time. Facebook rolled that stuff back long ago when it seemed there could be problems. And of course, the misrepresentation of how the data was acquired - because the story was tied into other entities who got 'special access'. The blending of those stories created a false narrative. To me - this is not about CA, it's about the press, clickbait, narrative management, PR and communications. Knowing roughly 'what's happening' and lining that up against the narrative, the glowing glam-shots of the whistleblower in the Guardian, Facebook's response, the various versions of 'outrage' among the politicos. And nary anyone seemed to be on top of the basic reality.
samstave 8y ago
That doesnt matter, that information wasnt as useful.... What matters is the intent and the methods. Its not overblown, its under corrected.
thinkersilver 8y ago
Bear in mind, I'm from the U.K so I don't have an American perspective on this. This is a bit closer to the truth when you read or listen to what analysis they were doing. I think what they were doing and it's association with Trump conflated how bad it was. I'd seen in some places that mining of Facebook data in the pre-trump era, was quite prolific by both parties.
- filoleg 8y ago
  Not just seems, it was a bragging point, because back then it was fine in the eyes of public, since the previous administration was a bit more likable. I recall an interview back from 2012, where our previous president was talking about it as an innovative technique for his campaign.[1] [1] https://www.technologyreview.com/s/509026/how-obamas-team-used-big-data-to-rally-voters/ https://www.technologyreview.com/s/509026/how-obamas-team-us...
crdoconnor 8y ago
This is similar to what I'd heard from people laid off by Cambridge analytica. Apparently credit rating agency data was much more useful for their purposes while Facebook data was pretty much useless.
- techntoke 8y ago
  Wtf? How did they get credit ratings and where was that reported? That sounds like a huge red flag.
  - cm2012 8y ago
    Companies like infogroup estimate credit score for all Americans. You can buy that + estimated income + addresses for every American for not that much money. This has nothing to do with social media. Credit card companies and other direct mail orgs have been buying and selling this stuff since the 80s.
    - spiderjerusalem 8y ago
      Do you have an estimate for how much 'not that much money' is?
      - cm2012 8y ago
        You can buy a person's data in bulk for a penny each.
  - nullwasamistake 8y ago
    Credit ratings are a dime a dozen when you're doing targeted ads. How do you think the 3 for profit ratings agencies make money? I get tons of mail telling me to consolidate my student loans, it's not random
cm2012 8y ago
I manage FB ads for a living. I can guarantee you the cambridge analytica data was close to useless (custom audiences almost always are compared to lookalikes. Especially a broad untargeted audience like that.) Much more impactful was organic (non paid ads) russian efforts.
- thesagan 8y ago
  And I'm still skeptical either were that effective. Domestic efforts go into the billions of dollars and involve entire sectors of the domestic economy.
  - cm2012 8y ago
    Probably. The biggest thing was not on social media at all - it was the drip releasing of the hacked email dumps, which controlled the news cycle for months and fed into the corrupt Hillary narrative.
    - Teever 8y ago
      I have a feeling that these things worked in synergy and would not have had much effect if deployed on their own. I'm not saying that the Russians or whoever influenced the election necessarily knew what they were doing from the outset but I think that they have stumbled on a combination that works or at least worked at that time. Which leads me to another interesting point: Will these techniques work again in the future? Will the application of them need to be tweaked for differences in our culture or in different localities? How so? It's all very interesting stuff. I wish there was more data.
      - zaroth 8y ago
        What do you mean by “these techniques”? I would say it boiled down to “telling lies that infatuate their target audience.” In this regard I’d say the collusion conspiracy kicked off by Steele has been significantly more efffective than any Russian effort leading up to the election. Mostly that’s because the Russians had some fake Facebook groups while Steele had the FBI and a large part of the American mass media all shilling for him.
JumpCrisscross 8y ago
> it sounds like Cambridge Analytica was overblown, in the sense that the information wasn’t as useful as claimed The consequences of the breach were overblown. Its scale and depth, however, were not.
- e40 8y ago
  The "overblown" part refers solely to the predictive powers of the dataset.
ErikAugust 8y ago
My conspiracy theory is that the whole Cambridge Analytica ordeal makes Facebook data appear to be valuable, which makes Facebook look valuable. And in reality, it is not nearly as valuable as the perception that is cultivated.
- JetSpiegel 8y ago
  This seems corroborated by the fact that Facebook's fine was incorporated as cost of doing business. Putting it in another way, CA was a false flag operation to keep Facebook's stock value?
  - propogandist 8y ago
    Yes. The masses (including investors) do not know much about facebook ad targeting, its effectiveness and RoI on social advertising.
e40 8y ago
I found this podcast fascinating, as it contradicted almost everything I had heard about CA. It seemed well researched, but now I'm wondering who is right? I'm leaning toward believing Alex Kogan, but honestly, I don't really know at this point. It's confusing.
cfarm 8y ago
Still should not have been accessible like this.
swarnie_ 8y ago
It became a nice little scapegoat for voters who lost on both sides of the pond.

sonnyblarney 8y ago

It was also overblown to the extent that there as a material 'scandal'. Most of the issue really had to do with the very public fact of Facebook's API's at the time. Facebook rolled that stuff back long ago when it seemed there could be problems. And of course, the misrepresentation of how the data was acquired - because the story was tied into other entities who got 'special access'. The blending of those stories created a false narrative. To me - this is not about CA, it's about the press, clickbait, narrative management, PR and communications. Knowing roughly 'what's happening' and lining that up against the narrative, the glowing glam-shots of the whistleblower in the Guardian, Facebook's response, the various versions of 'outrage' among the politicos. And nary anyone seemed to be on top of the basic reality.

samstave 8y ago

That doesnt matter, that information wasnt as useful.... What matters is the intent and the methods. Its not overblown, its under corrected.

thinkersilver 8y ago

Bear in mind, I'm from the U.K so I don't have an American perspective on this. This is a bit closer to the truth when you read or listen to what analysis they were doing. I think what they were doing and it's association with Trump conflated how bad it was. I'd seen in some places that mining of Facebook data in the pre-trump era, was quite prolific by both parties.

filoleg 8y ago
Not just seems, it was a bragging point, because back then it was fine in the eyes of public, since the previous administration was a bit more likable. I recall an interview back from 2012, where our previous president was talking about it as an innovative technique for his campaign.[1] [1] https://www.technologyreview.com/s/509026/how-obamas-team-used-big-data-to-rally-voters/ https://www.technologyreview.com/s/509026/how-obamas-team-us...

crdoconnor 8y ago

This is similar to what I'd heard from people laid off by Cambridge analytica. Apparently credit rating agency data was much more useful for their purposes while Facebook data was pretty much useless.

techntoke 8y ago
Wtf? How did they get credit ratings and where was that reported? That sounds like a huge red flag.
- cm2012 8y ago
  Companies like infogroup estimate credit score for all Americans. You can buy that + estimated income + addresses for every American for not that much money. This has nothing to do with social media. Credit card companies and other direct mail orgs have been buying and selling this stuff since the 80s.
  - spiderjerusalem 8y ago
    Do you have an estimate for how much 'not that much money' is?
    - cm2012 8y ago
      You can buy a person's data in bulk for a penny each.
- nullwasamistake 8y ago
  Credit ratings are a dime a dozen when you're doing targeted ads. How do you think the 3 for profit ratings agencies make money? I get tons of mail telling me to consolidate my student loans, it's not random

cm2012 8y ago

I manage FB ads for a living. I can guarantee you the cambridge analytica data was close to useless (custom audiences almost always are compared to lookalikes. Especially a broad untargeted audience like that.) Much more impactful was organic (non paid ads) russian efforts.

thesagan 8y ago
And I'm still skeptical either were that effective. Domestic efforts go into the billions of dollars and involve entire sectors of the domestic economy.
- cm2012 8y ago
  Probably. The biggest thing was not on social media at all - it was the drip releasing of the hacked email dumps, which controlled the news cycle for months and fed into the corrupt Hillary narrative.
  - Teever 8y ago
    I have a feeling that these things worked in synergy and would not have had much effect if deployed on their own. I'm not saying that the Russians or whoever influenced the election necessarily knew what they were doing from the outset but I think that they have stumbled on a combination that works or at least worked at that time. Which leads me to another interesting point: Will these techniques work again in the future? Will the application of them need to be tweaked for differences in our culture or in different localities? How so? It's all very interesting stuff. I wish there was more data.
    - zaroth 8y ago
      What do you mean by “these techniques”? I would say it boiled down to “telling lies that infatuate their target audience.” In this regard I’d say the collusion conspiracy kicked off by Steele has been significantly more efffective than any Russian effort leading up to the election. Mostly that’s because the Russians had some fake Facebook groups while Steele had the FBI and a large part of the American mass media all shilling for him.

JumpCrisscross 8y ago

> it sounds like Cambridge Analytica was overblown, in the sense that the information wasn’t as useful as claimed The consequences of the breach were overblown. Its scale and depth, however, were not.

e40 8y ago
The "overblown" part refers solely to the predictive powers of the dataset.

ErikAugust 8y ago

My conspiracy theory is that the whole Cambridge Analytica ordeal makes Facebook data appear to be valuable, which makes Facebook look valuable. And in reality, it is not nearly as valuable as the perception that is cultivated.

JetSpiegel 8y ago
This seems corroborated by the fact that Facebook's fine was incorporated as cost of doing business. Putting it in another way, CA was a false flag operation to keep Facebook's stock value?
- propogandist 8y ago
  Yes. The masses (including investors) do not know much about facebook ad targeting, its effectiveness and RoI on social advertising.

e40 8y ago

I found this podcast fascinating, as it contradicted almost everything I had heard about CA. It seemed well researched, but now I'm wondering who is right? I'm leaning toward believing Alex Kogan, but honestly, I don't really know at this point. It's confusing.

cfarm 8y ago

Still should not have been accessible like this.

swarnie_ 8y ago

It became a nice little scapegoat for voters who lost on both sides of the pond.