When we hear ‘Cambridge’, we consider a place of reverence. Cambridge, a place where science is academically pushed to new borders! It has been around for 733 years. In that time we saw Lord Byron working on Satires and poems. In 1812 Charles Babbage started the design of a calculating machine, he never finished it, but his work would later herald the modern computer. In 1903 Bertrand Russell publishes ‘Principles of Mathematics’ and ends up with being part of the ‘Principia Mathematica’ it takes people a decades to comprehend the genius and he ends up with a Nobel price. Other members will get similar laurels for working on the electron, X-ray diffraction, someone proves that vitamins are real and the atom gets split. there was Professor Stephen Hawking, who did have a sense of comedy (according to many sources), not very mobile, yet ends up giving us academic work on Black holes, the big bang theory (not the comedy) and gives us the a founding realisation on the origin of the universe and only recently do they were able to identify gene causing diabetes and high blood pressure. So we should see it as a place of academic goodness. Yet when you take ‘Cambridge’ and you add ‘Analytica’ you get a synonym for ‘Despicable Sewage‘.
So as we are treated just over an hour ago to ‘Facebook’s Mark Zuckerberg finally addresses Cambridge Analytica scandal‘ (at https://www.theguardian.com/technology/2018/mar/21/mark-zuckerberg-response-facebook-cambridge-analytica) we can clearly see that things are escalating as Mr Zuckerberg himself is taken off the moth balls to remedy the situation. So when I see his response ‘we made mistakes‘, my initial response is ‘You think?‘
When we are treated to “The Facebook CEO broke his five-day silence on the scandal that has enveloped his company this week in a Facebook post acknowledging that the policies that allowed the misuse of data were “a breach of trust between Facebook and the people who share their data with us and expect us to protect it”” my initial worry is that he does not comprehend the scale of the issue. It is not merely the misuse of data, basically personal data of 50 million people, a lot of data on these people is now out in the open. When you have the data of 14% of your population you have the means to forecast, the options to set the marketing push on a national level. That amount of data would allow places like Walmart to set the need to satisfy 90% of the population need and cut out the loss making products overnight. You see, when you take the concept of a good article, a average article and a bad article, we often get all the good articles and a chunk of average articles. This is the risk the business has, they all have it and we can predict this to some extent. Now we get more data and now with that data we see a group of people that are classified for a certain category as ‘Not caring’, they have no interest at all. Knowing this allows for the setting of a ‘true view’ on the articles so we get a sharper view, we take the population, we take out the non-carers of that product, and suddenly we end up with a list of the products that are all classified as good.
Now how does that work?
You see, sometimes we are driven by internal motives, motives we do not tell Walmart, but we might tell others on social media. Now consider for example that a Catholic will never buy a certain brand. A naturist will never buy certain chemicals and a tech-lover will never buy certain brands. There are dozens of these indicators and Walmart, if they had that data can now see a pattern, even if they only have the 14% view, the pattern once seen can lead to a national view. As a wild example I give you: ‘A Catholic techie will always buy a Manfrotto camera stand‘. So now we have a specific product that would do really well in Rhode Island, Connecticut and Massachusetts. So not only can it decide to dump the inferior camera stands in those places, it could essentially also raise the Manfrotto price by $2, so less overhead and better profits. This is merely an example, but the pattern is clear and as places like Walmart have such data they can now directly target their audience and streamline what they carry per location. So not only do they get a better business setting, by marketing directly to certain groups they get a much better result on the same marketing cost. So their marketing costs remained the same whilst getting up to a speculated 30% of better results.
This is a given setting in analytics (and Market Research). It has existed for decades and Mark Zuckerberg is a clever boy, so he knew this. The setting as shown in the Guardian is debatable at this point. You see, debatable because Mark Zuckerberg knows the value of data, there is no way that he does not know that. So the last thing he wanted to do was hand out data, lose control of the treasury. He lost control as the data is out there now, and as the source has been shared for what I believe to be at least three times over, that data is now no longer containable. That can now be seen as a direct loss for Facebook.
In equal measure we need to look at “We know that this was a major violation of peoples’ trust, and I deeply regret that we didn’t do enough to deal with it”, a quote that came from Sheryl Sandberg. You see, I think that the matter is more serious and more dangerous. We see that when we realise that ‘we didn’t do enough to deal with it‘, there is a data quality loss, a data containment loss and a lack of technological oversight. This is not a new given and even as Cambridge Analytica took it to a much larger setting, they were not alone. I myself almost tried a game once, yet when I saw it wanted my ‘religious preference‘ I decided to have an issue with a game firm that is concerned with my religion. I don’t have any, but that had absolutely no bearing on the game. That made me suspicious and I decided not to install the game. There has been a flaw for the longest of times. That flaw goes all the way back to Zynga’s Farmville. When they started to demanded ‘gifts from friends‘ to progress to some extent in the game, it was not a novel thing (well it was), it was a marketing setting that either you pay for the next item (with buyable currency), or you get your friends to play the game and give it to you, so we saw groups of people all linking, whilst their only link was the social setting of one game and Zynga ended up with the data (to some extent). That requirement is not what I see as ‘social growth’, it is in its foundation a dangerous place because it allows paedophiles access to younger players, it allows white supremacists to hide in a social flock whilst the others in the flock had no idea that the herd is not just made up from sheep, it also contains wolves and other undesirables. The problem is that as long as nothing happened no-one would care and that has been a dangerous game to play. Facebook loved the concept because it grew communities beyond their wildest dreams, but it also gave us groups where we still needed to be careful what data got out, yet the people at large are not careful with their social data. That has been seen since 2011 as Prostitutes were found by several media publications to use Facebook as a customer recruitment system. Now, I don’t care what these ladies do, yet as we have seen that recruiters and HR are using Facebook more and more to ‘judge’ potential employees (and one should never talk to a ‘lady of the night’ in social circles), we see that Facebook has become a monster of abuse and that monster is valued for data, so as more and more data is added, more and more people end up getting wrongfully tainted in a colour that was never them.
So when we see “The CEO also pledged to investigate and audit apps that accessed large amounts of data from Facebook users prior to changes in its platform in 2014, and said that it will inform users if their personally identifiable information was misused by app developers“, we need to realise that the foundation of Facebook apps is a much larger problem, it is not merely about the data they can access, it is the issue we see when the app data itself is open to mining. You see it is not merely “Facebook will investigate all apps that have access to large data and ban developers that misuse identifiable information“, how about apps that merely collect a small amount of data. Now consider that they link the use of apps (like for example Farmville, the Pioneer trail and Cafe World). Now let’s be clear, I am not accusing Zynga of doing anything wrong or illegal. But those three apps allow for ‘free’ currency, when you hit a target in the other game, people start to get very motivated to play 3+ games from the same makers, as it allows for that currency that is not usually free. Consider that each app has 5 demographics and perhaps 3-5 additional stats and these three apps all have 3-5 different stats. So as hundreds of thousands are playing all three apps, the developer suddenly ends up with a much larger pool of data than ever before. Now I use Zynga as it also has real-money gambling games. Now consider that they now have more and more markers on people who gamble. It is the wet dream of any Las Vegas entrepreneur to get that much data on their users, a way to classify those who are more likely to spend more on gambling. This was a setting that has been known for a long time and there is no way to tell how far people ended up being pushed into gambling. I have seen and learned that greed is eternal, so in that regard there is all likelihood that Mark Zuckerberg had to be aware to a much larger extent and that the mistakes made were a lot more than ‘excusable’ and it is one that cannot be solved through apologies and better oversight. Because when we cut those developers off from the data, what are the chances that 70% moves elsewhere? Data was the treasure trove and too many have been dipping their toes in the water. The damage is much larger and even as Cambridge Analytica made it visible to the masses, the issue has been there for a lot longer, the question becomes who properly looked at it. Also consider that games like Farmville had well over 60 million users every month, so how much data made it out of Facebook?
I reckon that no one will actually know that part, but the issue is also how this simple given remained off the radar of so many for so long. I wrote about the sharing of data as early 2013, sharing at the same time my thoughts on how all the NSA data issues were merely hypocrite. Well, now that the fence is gone, good luck containing the sheep, because I expect to see a lot more ‘revelations’ over the coming months.
Lawrence van Rijn - Law Lord to be 