Cambridge AnalyticaSeptember 2nd, 2019
I recently reviewed the video The Great Hack from Netflix.
Upon looking to cite the link above as a source, I stumbled upon a top Google result (currently second only to the Netflix trailer page itself) called
The British consulting firm didn’t steal the 2016 election, and it couldn’t have if it tried.
Since this view aligned with my thinking as well, I figured I could save a lot of time writing this and just agree with the above. And while I agree with the fact that CA did not steal the 2016 election, I feel like I have more to add with specific examples from the movie being fresh in my mind.
The Netflix documentary eludes to CA having limitless access to Facebook's ad platform. This is definitely not the case. While I do expect Facebook to favour advertisers that do spend $1M USD/day in ads with additional targeting features, the movie does not explain how the ad tools Facebook provided granted CA the capability to utilise their audience in combination with 5000 data points/person.
The biggest crime I saw in the movie, was a data breach on Facebook which was Facebook's own fault for having poor security policies in place which caused a breach of friend information unknowingly. Facebook did take measures to resolve this, but this data got indexed. That is, copies of this data were made. And so, with this data in high demand, the problem of suppressing it would be equivalent to that of trying to remove a celebrity sex tape off the internet.
Cambridge Analytica could have in fact deleted the information as requested by Facebook. Brittany Kaiser's claim that in an email shortly after one of the high-ups at CA mentioned they could "use Facebook likes" does not concretely imply that these were "likes" which were held onto despite the request for deletion. I am saying, CA could have very well deleted the data as requested and still completely legitimately outlined in an email that "likes" could be used as a targeting tactic.
If somebody can prove that what was in fact said was along the lines of "we can use the stolen likes we were supposed to delete and didn't to target..." then that is good evidence to confirm Brittany's claims, but that is not what was presented. Brittany also provided questions for Senators to ask Zuck. *I wondered where they got those questions* And her question, with regards to how much data is used to monetise Facebook and them joking about Zuck being forced to answer "all of it". So what? Would you want the answer to be "half of it"? In fact, such an answer, when put on the spot might be very difficult for Zuck to quantify on the spot. I am sure it isn't "all of it". If it was "all of it" that would imply the post I made in 2008 requesting a good plumber which has since been made private could be used to target me with to prospective plumbers. If Facebook was using all of the data, then plumbers would be targeting me based on the fact that I needed a plumber 11 years ago. It doesn't make sense. Facebook needs to digest the data down to a subset of classification in order for it to be useful to advertisers. And the more recent the data, the more relevant it would be. It's not a clear cut answer and it is a stupid question to ask designed to make Zuck look like a greedy villain. In fact, the limit of $1M/day ad spend on Facebook is really the main constraint in terms of access to Facebook data.
As a Facebook advertiser and app developer myself, I remember Facebook adding new policies to correct the massive data breach they had. One such policy is that user ids are no longer universal across Facebook apps, but specific to apps themselves so that data cannot be mined from multiple sources and easily joined together.
The main data breach issue was fixed before the election and app developers were expected to delete user data from their databases upon request from Facebook or Facebook users through app removal requests.
The second crime I guess you could say is that CA did not release the user data upon request as required by law. And they plead guilty for that. Why they didn't do it? Your guess is as good as mine. But generally, in my experience the reason people do things is because they are priority and the reason they don't do things is because they are not priority. So my conclusion is that it was not priority to do that. With all of the attention CA received after this, I expect they could have been potentially flooded with other requests which could have made this single issue a needle in a haystack. That is, CA may have failed to meet many requests. But the way the video is narrated, it makes you think that CA had nothing better to do than scour their data for a single individual, albeit a legal requirement, but with all CA was dealing with, I am not surprised it was not done.
Also, there are many tools and data that I have built which I would be the only person with access to and capable of retrieving and reporting on. If Brittany herself in the video (a CA ex-employee with a conscience) was the one who had the level of access required to retrieve this, then her departure perfectly explains why it was not achievable.
As the article mentions, you should expect an advertising company to embellish their ability and capabilities. You can honestly say that you have data points on 57M people, and conveniently leave out how many of these data points are useless and/or potentially obsolete. You have to expect these companies to use big numbers to back their claims.
Getting back to the $1M USD/day claim that the Trump campaign was spending through CA, with the "majority going to Facebook". This has huge impact no matter what side you are bidding on.
I think what bugs me the most about the attacks on CA and Facebook is the hypocrisy of it all. If Hillary Clinton had won the election, Cambridge Analytica would be the losing agency. They would have been left alone for certain as there is nothing wrong with legally helping a politician run ad campaigns. And if this were the case, would Hillary's ad company undergo the same scrutiny? I think not.
In my opinion, television backed Hillary and Social Media backed Trump. Television lost and is using their legacy media to create their own fear campaign of Facebook.
The movie mentions at least 3 times "weapons grade advertising". Which is funny because I feel like all advertising is a weapon. The truth is television is bringing a knife to a gun fight. Television has a complete inability to target gender, age, race, religion so they have to create content that targets those people and sell ads in that content. Ads that are designed to be difficult to avoid and force you to sit through for several minutes. Like hockey? You will love this beer. Like Oprah? Here are the tampons you should be using.
With television, viewers have to first be convinced to view. Facebook offers many things that you cannot get from television and in the end, both compete for your line of sight.
Facebook is far better at manipulating you with ads, then television is. And that is a major theme the of the movie. They use FUD (fear/uncertainty/doubt) tactics to make you fear how powerful and targeted the Facebook advertising platform is by making you feel like they have stolen your data without your permission and given it away recklessly. You should fear that data has grown to an industry bigger than oil and it is not up to you to carefully monitor what you put on the internet, but instead it should be in Facebook's hands to forcibly delete any trace of you which originated from Facebook upon your request. I know better. You can't expect Facebook to be 100% effective in deletion. That is a best case scenario. If you post a titty pick and it gets deleted off of Facebook, do you really think a bunch of guys didn't right click save as that shit? Do you think asking Facebook to remove it deletes it off all your friends hard drives? Facebook can only do it's best to delete your stuff. If it really up to the users to be careful about what they post.
The privacy controls at Facebook are actually pretty good. Assuming no breach, you can limit exactly who you want to have access to your content. And you choose the places where you login with Facebook login. And you choose the third party apps you install and link to Facebook. Each one of those companies provide a list of permission scopes which they require from you in exchange for app use. These apps are now carefully monitored by Facebook to ensure permission abuse is minimised. I don't believe for a second Facebook is not doing everything they can to keep your data private while giving you the best user experience they can. And yes, they did design the site game like so that endorphins are regularly released in your brain from your friends liking and approving your content. The result is a very good system to allow advertisers to target you by age, gender, location, interests etc. etc.
Would you rather see an ad where (you are a man) and the ad is for tampons, or a helpful suggestion for a gift you may get your wife for her birthday next week? This example encapsulates why television is a stupid place to advertise and Facebook should legitimately deserve more ad $.
And while CA claims to be able to take 5000 data points and filter these down with special algorithms which determine who is convertible voter, I challenge that $1M/day budget reaches far beyond those convertible only voters. If you feel outrage by the sheer number of data points a company can collect on you, you really shouldn't. Because ultimately each advertiser wants you to buy what they are selling. And all those data points combined are not as good as the a single data point generated with a survey such as:
Who are you voting for:
- Hillary Clinton
- Donald Trump
If the big problem is that CA was too effective in their advertising and you discount all the work Trump did in campaigning effectively himself and ignore the ad spending which increased line of sight, then you are left with a campaign message "Crooked Hillary" and the oo's make handcuffs.
The real issue in my mind is that "Crooked Hillary" was allowed at all. What does "Crooked Hillary" make you think? Fear? Uncertainty? Doubt? That's what I get.
Perhaps, instead of putting so much effort into combating Facebook and claiming they are guilty of wrongdoing, the efforts are better served in calling out media campaigns with the intention of spreading FUD. FUD is an extremely effective media tactic to manipulate perception through negative emotion.
The definition of FUD from Wikipedia states it is based on a disinformation campaign. While this may be true in many cases, when a person is presented with binary options to support or not, using fear to scare somebody away from one side is a dangerous weapon. And the weapon is more deadly based on reach regardless of whether it is founded in truth or not.
The weapon's used by CA were not targeting individuals. The weapons they used were effective data collection and historic experience and success with FUD to create viral campaigns which drew support on a platform which gave people multiple reasons to fear, doubt or be otherwise uncertain about one candidate.
If Trump lost it would have been because America could see through the tactic of "Crooked Hillary" and fail to believe it. And while many took this stance, the dirt on Hillary through Wikileaks, her lack of security precautions and following protocol which lead to a data breach which exposed DNC emails, her husbands own sexual deviance, ties to Jeffrey Epstein, and Mena Arkansas add up to enough doubt about Hillary's ethics that it swayed me.
I ate up all the Hillary stories. Why did I do that? Because I abandoned cable TV years ago. They weren't my line of site any more. They couldn't convince me that Hillary was leading in the polls or that Hillary would certainly be the next president. I thought she was crooked. And with Bill's ties to Jeffrey Epstein, the "Lolita Express" and the respect I have lost for Hillary not following security protocol, I do have little sympathy for Hillary.
In fact, when you compare Hillary's breach of protocol at the government level, to Facebook's data breach, I think they are both human mistakes but Hillary's was much more deliberate as it required extra effort whereas Facebook's breach was caused by lack of effort.
The DNC picked a shitty candidate. Maybe colluding with CNN regarding debate questions is a common thing. But bringing that knowledge to the forefront on Facebook shows Hillary as a cheater. More doubt.
In conclusion, I think it is a good thing that the Netflix video was made. I think that vilifying tactics that increase and target reach more effeictively is dumb. Focus should be put on regulating the message. Campaigns which discourage voting should be outright banned from all media for starters. Each message should undergo disinformation scrutiny.
Should Trump have been allowed to use the term "Crooked Hillary"? I don't have the answer to that. But it is worth discussing. If somebody speaks truth about somebody else in a negative tone, is it fair game? I think yes. But if yes, then we become vulnerable to differences in opinion and realities the way some perceive it. So maybe we say no? But if we say "no", then how do we point out when somebody is corrupt?
I think campaigning needs a good hard look at the ethics behind messages they send. It shouldn't be allowed for Bernie Sanders to call Donald Trump a racist as if it were known fact. He should have to supply a citation. I think that any smear needs to be based on a predated cite-able source which is challenge-able by any media source or maybe even any citizen. If Bernie Sanders cannot produce a source which essentially proves that Donald Trump is a racist, he should have to make a public retraction. Retractions would work against candidates.
Did Trump cheat? In my opinion no. Should he have won the election? Was he the best candidate? I agree with those who say "no". But I think Trump should not have won the election because of his dishonesty. I think the fact checking and scrutinising of every argument he says should be debatable on both sides.
If such a system existed where you could consult how much inaccurate information politicians used in their campaigns, people could at least have the option in determining whether they factor these inaccuracies into their decision making process.
If such a system were web driven, and a politician feels they are able to cite a source that everybody might think is weak or questionable, then tools can be used to up/down vote these sources so that people can review their weakest source citations.
Maybe each claim made by a candidate defaults to retracted. It is up to the candidates to backup their claims with citations. In the end, you could have a summary or several summaries which can be graphically depicted which shows how good or bad each is at telling the truth.
I believe that I may use such a system to determine who I vote for. And in doing so, if I find myself in a position where I do not know who to vote for, I can review the worst citations from each candidate and make the decision on whose I feel is the most ethical. Because at the end of the day, for the world to really succeed, you need to have the most ethical people in positions of power.
Software Developer always striving to be better. Learn from others' mistakes, learn by doing, fail fast, maximize productivity, and really think hard about good defaults. Computer developers have the power to add an entire infinite dimension with a single Int (or maybe BigInt). The least we can do with that power is be creative.