Distant Innocence: The Zuckerberg Factor

Facebook. You might have heard about it, the "move fast and break things, "bring the world closer together" folks. Home for a gagillion cat videos, cute baby pix, political rants, braggadocio, Khalil Gibran quotes, birthday wishes, bike ride photos, fake accounts, hoaxes - and ads and ads and more ads.

Facebook knows everything about you ... because you've told them. You've told them explicitly on your profile page, you've told them implicitly through your posts, your likes, your browsing history, your purchases online. The thing is .. this data gets leveraged for all kinds of purposes. Go to Amazon and search for that a paper shredder (you know, to protect your privacy) and the next time you bring up Facebook, there's an ad for paper shredders. Facebook automatically acquires data from Amazon, Google, Bing and on and on. Over 4 petabytes (see chart below) of data get uploaded to Facebook daily.

Every day:

500 million tweets are sent (some not from Washington)
65 billion WhatsApp messages
294 billion emails
5 billion searches
4 terabytes of data from each connected car

Think of a byte as 1 character

This is the world of "Big Data" and data analytics. Everything you do, everywhere you go, everything you buy gets tracked. "IBM estimates that every day 2.5 quintillion bytes of data are created – so much that 90% of the data in the world today has been created in the last two years." It is mind-boggling. By 2025 the estimate is 463 exabytes (see chart) of data will be created each day.

So what? The problem is that with all this information available, it and you can be manipulated and sad to say not always is for altruistic purposes. Most retailers have robust data analytics capabilities, utilizing their customer data, sales data, etc. How do you suppose it is that Target could know your wife is pregnant before you, her husband? Data analytics.

How is it that small market baseball like Oakland can be competitive against the $250 million Yankee payroll. Data analytics. Oakland finds players which other teams overlook or undervalue.

Big data / data analytics have beneficial, perhaps mostly beneficial intent.

Medical: evaluate symptoms and identify many diseases at the early stages

Ecommerce: evaluate customers behavior and suggests similar products.

Disaster Management: identify the potential disasters by evaluating temperature, water level, wind pressure, and other related factors.

Security: detect fraudulent activities / identify the possible threatening situations

Agriculture: take data from the past years and can suggest the pesticides that work best under certain conditions.

The capabilities have become so sophisticated it is hard to know what to believe, so generally people retreat into their clans and accept everything that fits and ignore anything that challenges that world view.

But there is also a large set of dubious data analyses applications. One of the most enthusiastic adopters of big data & analytics is political campaigns, leveraging Facebook's "micro-targeting" capability. The Trump campaign recognized the power of The Zuckerberg Factor early on, purchasing 6 million FB adds in the run-up to the 2016 election. It is likely you saw but a few of them. The targeting is so sophisticated that ads can go to, for example, "suburban hockey moms in Buffalo". Or more disturbing, racists in Toledo, with Facebook inferring from peoples posts and comments, etc. who might fit that profile. Which Catholics in Wisconsin have attended Mass 3 times in the past 3 months? Hmm, why do you ask? Doesn't matter. Facebook really has no rules when it comes to political advertising. Only rarely would they reject a political ad. They would not allow advertising a false election date, but that's about it. Under duress, they recently tagged an altered video as "partially false". It's just fine for politicians to lie on Facebook. There is no fact-checking.

And there are a wealth of FB capabilities you probably are unaware of. For example, Facebook can produce "look-alike lists". You provide a custom list of people and minimal demographics (FB can easily fill in the holes) and FB will give you another list of people who match that profile. So - provide a list of 250,000 supporters and FB will give you another list of 250,000 potential supporters.

The new language of these political machinations is ominous (if you really want to nerd-out, links are provided into this new world of analytics arcana)