Would we have lost Bowie to the algorithm? 

Opinion

WPP’s David Wilding explores whether, if Bowie had come of age in the mid-2020s rather than the mid-1960s, we would have lost his genius to his feed and the algorithm.

In 1999, David Bowie gave a famous interview to the BBC in which he said:

What the internet is going to do to society, both good and bad, is unimaginable. We are on the cusp of something exhilarating and terrifying… the context and the state of content is going to be so different to anything that we can really envisage at the moment… the interplay between the user and provider is going to crush our ideas about what mediums are all about.” 

He wasn’t wrong, was he? 

The interview features in the current David Bowie: You are Not Alone exhibition at The Lightroom in Kings Cross, London. Featuring Bowie himself as its sole voice, it’s a multimedia showcase of some of his most iconic performances but also an intimate self-reflection. I went last week and have been thinking about it ever since. 

Two big themes of the exhibition are Bowie’s voracious cultural appetite and his insatiable curiosity. In his own words, Bowie was “a very curious and enthusiastic person”, drawing inspiration from a diverse range of sources and constantly exploring and developing different interests spanning music, literature and the arts. 

Bowie talks about how:

I would see every movie, I’d go to the theatre as often as I could afford to. I would often buy books that were far above my head, but actually these things kind of made their mark because I found out that I’d actually end up reading them, and then I would be exploring territory that I never thought I would get into”. 

In the few days since seeing the exhibition, I have found myself pondering: if Bowie had come of age in the mid-2020s instead of the mid-1960s, would we have lost his genius to his feed and to the algorithm? Would he have been exposed to such a broad range of seemingly random stimuli, or would his algorithmic feed serve up more of the same?

Now, we’re talking about David Bowie here, so a fair and reasonable answer to these questions might simply be “yes, he was David Bowie”. But for everyone else, the evidence suggests that the odds of finding variety, the unknown and serendipity in our algorithmic social feeds are stacked against us. 

The dark arts of the algorithms

In his book Robin Hood Math, Noah Giansiracusa outlines how the algorithms used by social media platforms are designed to surface the posts that people are most likely to engage with.

There are several forms of engagement available to monitor (e.g. reshares, likes, reactions, completed views, comments), and the algorithms weigh these different types according to the strength of the predictive signal and then use a weighted sum to provide an overall measure of predicted engagement.

Whichever content scores highest on predicted engagement is then served next. Each time this happens, the algorithm gets a stronger and stronger signal of what we are likely to engage with. 

In practical terms, this means that if you tend to engage with food posts, the algorithm will, over time, increase your estimated engagement probability for food posts.

What this means is that – theoretically – any of us can ‘train’ our algorithm to some extent. If we want more food content, we should like, love, comment on, and share it. If we don’t, then we shouldn’t engage with it. 

And none of this is a secret. Instagram advises people searching for information on how to control their algorithm to view a summary of the topics Instagram thinks we’re into, then adjust their feed to see more or less of those topics. You could liken it to gardening. Plant the seeds you want to see bloom over time, and prune back the rest. 

The problem is that most of us aren’t gardeners. And even if we are, what if we don’t know what we want to see in the garden but are open to inspiration?

For most people, most of the time, proactive control of our algorithms isn’t something we’re consciously discussing or doing. It’s very far from the default, so most people don’t do it.

And the negative impact of this is increasingly being understood. If your social media algorithm isn’t a healthy one, it can be very hard to break the cycle. Doomscrolling very quickly begets more doomscrolling. Misinformation and disinformation lead to more engagement, but more disinformation in your feed. 

The fightback is starting to take shape

ScrollAware is doing some excellent thinking in this space (the Media Leader podcast interview with Jess Butcher is well worth listening to). Pinterest has just released a beautiful provocation that asks, “How did people (used to) know they liked something if the thing that they liked hadn’t got any likes?”

Snapchat made the case at “Snap Forward” last week for “the great social media reset” and claimed that “The next era of social media is already here” – and that it is quieter, closer and built on real connection. 

And let’s not write off the algos completely. Just because they are currently optimised for engagement – and therefore to maximise the time we spend in our feeds – doesn’t mean that this always has to be the case. You can optimise an algorithm for anything; there’s no technical reason why they couldn’t be optimised for positive discovery, serendipity and different perspectives. 

In the spirit of Bowie, I’m highly curious about how we do this. What do you think? 

David Wilding is EVP strategy at WPP Media and writes a monthly column for The Media Leader.