In browsing through my Facebook and Twitter feeds today, and reflecting on the relation between power and algorithms (Beer 2017), a friend posted on Facebook a Guardian article asking ‘Why don’t we treat the climate crisis with the same urgency as coronavirus?’
Immediately below, a post was displayed from another friend (from different circles) linking to an article from Business Insider about the importance of hand washing and the coronavirus.
It’s difficult to tell why the algorithm placed these posts next to each other – whether there was a connection between the two, or whether both were deemed independently likely to exhibit some kind of ‘positive reaction’ (or reaction of another sort) from me. However, it did make me reflect on how initially – to a degree – these posts felt (at least to me) “pitched” against/at odds with one another (by being placed in close proximity) and exhibited initial reactions that I might not have otherwise felt had I seen the posts independently.
Turning my attention to Twitter, this kind of algorithmic restructuring of timelines, at times using “deep learning”, initially caused controversy and an #RIPTwitter “campaign”.
@mjahr talked about these early efforts on the Twitter blog…
‘…when you open Twitter after being away for a while, the Tweets you’re most likely to care about will appear at the top of your timeline…’
(@mjahr on Twitter blog, 2016)
…and went on to praise their “success”…
‘We’ve already seen that people who use this new feature tend to Retweet and Tweet more, creating more live commentary and conversations, which is great for everyone.’
(@mjahr on Twitter blog, 2016)
Furthermore, in this HootSuite blog post, criticism of the Twitter algorithm was downplayed somewhat, stating that ‘the algorithm drove more engagement from users’.
Yet what does ‘engagement’ mean, how do they know what we ‘care about’ and how can they possibly prove that it is ‘great for everyone’? Driving traffic and increasing usage may be ‘great’ for both Twitter and HootSuite – companies who attempt to derive profit in one way or another from such usage – however, this kind of biased uncritical language is perhaps all too common in some circles (such as those who are profiting through advertising or sale of associated products).
Much has been written about the questions and issues that algorithms such as these raise, not least the relationship between political rhetoric and social media; for example, by Oliver (2020). I continue to read, explore and critically reflect, particularly pondering over what this might mean in an educational context (such as our own use of Twitter during this course)…