Facebook 10 videos – how brains think (but computers can’t)

frog picture

Facebook thinks this is a significant moment in my life

I saw my Facebook 10 video today, their 10th anniversary “gift” to members.  It’s an algorithmically generated video montage of posts and pictures from your timeline, in my case going back 6 years.

It was dreadful.  But a useful example to think about how important memory formation is to your understanding of life.

Facebook 1o – edited highlights of your Facebook life?

The system works by pulling out the pictures that got the most shares, the posts that got the most likes and so on.  However, in my case, it was mostly an aggregate of all the insignificant moments I’d forgotten (because they were eminently forgettable) or worse, reminders of places I’d been or people I’ve since lost touch with, for good reason – including, the wedding of a former business associate who presided over a dodgy share deal that led to me breaking off all contact with him for the sake of my professional reputation.  I’d forgotten I’d ever posted those pics.


What did it miss?

Let’s see… my marriage, the birth of both my children, my family, my in-laws, my home, my pets, in fact, all the most significant life events I can think of.  Which got me thinking.  You see, how my data footprint defines me and how I define myself are very different things.  More than that, how my Facebook posts get shared within my network of Facebook friends isn’t reflective of the importance of those posts when viewed from the perspective of the person who made them.


What did I learn?

Actually it gives a very interesting insight into how our emotional, unconscious minds help us rationalise the irrational.  Facebook is just an online storage mechanism for your words and pictures, with a bit of email and chat messaging stirred into the mix.  Other people intersect with your data on that functional, technical basis.  What happens then, in reality, is a set of data is created that records those real-time intersections between one set of binary numbers and another, generating a meta data set that describes the interaction of a network of pictures and posts from a group of people.

It has no insight.  The data doesn’t know who you are, or what you care about – just spots numbers that define how large an interaction was.  It doesn’t make the interaction itself important.  It’s not a score, or a result or any other differentiator that let’s you make a qualitative judgement about the event in question.

For example, in my timeline there’s a picture of a frog which I took, some years back, sitting in my pond.  It was quite a good picture of a frog, not great, a bit blurry, just a snap really. I’m not proud of it, I’d forgotten it.  It got lots of likes.  It is therefore in my Facebook highlights video.  When I look back upon the last six years of my life, which include births, deaths, success and failure, friendships made and lost, business booms and slumps, tears of joy and laughter (etc etc blah blah) – the frog picture is, unsurprisingly, nothing more than a picture of a fucking frog.

My experience of the emotional and social context within which life events occur, however, naturally makes me rationalise the events in my life into an ordered list of items, ranked in importance, and memorability.  It’s a fluid, ever changing list where the list items, depending on my mood or subsequent events that change your perspective on the past, change on a more or less daily basis.

The frog was forgotten to me but indelibly ranked as a significant life moment by my data.


The memory transformation trick:

That clash, between data and perception, illustrates a lot about how we think and form memories.  Computers track only what computers can track – events defined by numbers.  They might be very complex calculations, which treat the numbers as sets of interdependent variables which create a fluid result that might change as the numbers within the calculation change, which is very useful in things like the system that stops your car from skidding, or the system that keeps a plane on autopilot through unpredictable weather patterns.  But it’s not thinking. It’s measurement.  A system like that measures sets of variables (wheel speed, spin, altitude, pressure, gyroscopic acceleration etc.) and uses those different variables to adjust mechanical settings to provide a certain kind of result.  It’s very, very fast counting and button pushing.  It’s not thinking.

We, on the other hand, are thinking all the time.  We can’t count as fast as a computer, or perform tasks as quickly, but we can do something much,  more complex.  We experience events and record them as memories, however, as we record more memories and analyse the decisions, choices and events they relate to with hindsight, something remarkable happens – the memories change from being one type of thing to being quite another.

For example,  a good day can become a bad day in hindsight, like wedding day pictures, viewed in the midst of a messy divorce.  A celebration of life can become a poignant reminder of a tragic death,  like birthday pictures of someone a year before they died suddenly of cancer.  An insignificant, forgettable shot of your workmates down the bar can become deeply significant if it was the first time you met the person you build a multi-million dollar empire with, or (in my case) a happy shot of colleagues at a wedding can become a rather uncomfortable reminder that whilst the picture was being taken, the smiling guy next to me was signing papers without my knowledge that, had I known, would have made standing next to each other (smiling away at the cameraman) unlikely.

To the algorithm that picks out your significant life moments from your Facebook data, that transformative effect of memory is, of course, impossible.  The frog picture is important because important is defined as a variable.  It’s a fixed point in the algorithmic universe.  It’s like gravity or the speed of light, something that defines the nature of existence.

To the algorithm,  if you post a picture, or a comment, or anything else that meets the criteria, then it’s important.  In reality, a Facebook event might be popular because of the time it was posted relative to the time most of your friends were online, or the popularity of a particular genre of content, or any number of random variables you may or may not be aware of. Those are circumstantial, unpredictable, irrelevant criteria for assigning significance to an event as far as your ‘person’ is concerned. It’s just like people forming opinions, something quite outside your experience of life or control of events. It’s says nothing about what you think or who you are. It says nothing about your life. It’s a word from the story, it’s not part of a narrative.


The future is still algorithmic.

Of course it is.  You see, humans are, regardless of what we think, neural network computers (your brain).  Neural networks are very complex, they facilitate intelligence, the essence of which is expressed by memory formation and change, the ability for information to change radically as it grows around a specific context.  Facebook isn’t a neural net… yet.  Artificial intelligence is advancing all the time, and no doubt, a few years down the line they’ll start deploying systems that can moderate the importance of different sets of numbers to reflect the way your brain sees your memories, rather than the clicks, likes, shares (and anything else you might measure) today.

To put that into perspective, I have “liked” the Facebook page for Redbull the energy drink.  I “liked” it (and do still like it) because it’s full of kickass videos and pictures of racing cars, BMXs, stunt pilots and extreme sports stuff.  However, the drink Redbull, on the other hand, tastes like piss and makes me feel like I’m going to have an aneurism.  I don’t like Redbull.  I like Redbull’s Facebook page.  I don’t do extreme sports or BMX biking, but I like watching it.  How much?  Not that much.  In fact, as my memory processes the events that led to me “liking” the Redbull Facebook page, I realise I’ve only ever looked at it once or twice in the last five years.

Come to think of it, I don’t really like it at all, I saw it and thought it was quite cool, at the time, but don’t really give a shit either way.

Brains are tricky sometimes.  Frogs too.