Against A Better Algorithm
We don't need Algorithms to save the internet. We need incentives.
The fatal flaw in Jack Conte's otherwise uplifting opinion piece on Social Media Algorithms for the NYT, seems that he's maintaining the incentive structures and gatekeeper architecture of the companies he singles out as bad.
The Patreon founder enthusiastically addresses the problem of online social spaces that are no longer social - optimised for attention, not community or creative sharing. His solution to sites like Facebook, accused of worsening the mental health of its users (https://lnkd.in/ediM4zj6), Instagram, YouTube and others is to build a better algorithm. This, he proposes, might combat the steady worsening of our online experiences.
He's right that a "better algorithm" could be optimised for long term relationships, artistic reward and control, but the reason Facebook et. al. have moved away from that ideal over time is not that they are inherently evil (well, maybe a little bit...) but instead their structure rewards only one direction of travel. Their survival depends on (a) being the gatekeeper for user and artist created media and (b) directing users to the most successful (most profitable) content.
Those incentives nudge the algorithm constantly towards attention manipulation.
And platforms like Patreon have exactly the same incentives.
You can see it in the opinion piece. He's not arguing for giving artists ownership of their relationships with fans, but for Patreon to be benevolent shop keepers on their behalf, "trust us, we'll be kind".
"Our algorithm will be better"... why? And for how long?
Remember Google's original motto was "Do no evil".. How's that working out?
The problem here is the incentive structure, not the technical details of how content is delivered to you. Technology in this area is a mainly solved problem. We know how to serve, search, rank, charge and all the other parts of a great global scale internet.
The difficult part is the social part - the incentive structure that rewards good behaviour and beneficial innovations and that prevents bad actors monopolising a space.
We can certainly attack that social problem, and develop sites and protocols that offer a meaningful change. But merely delivering a "better algorithm" is a sugar pill that hides the bitter taste of the same old medicine.
The uncertain and challenging part is investing in new platforms designed from the ground up to incentivise a better relationship with the vast sea of amazing artists, journalists, makers and hobbyists that generate the stuff that brought us online in the first place. In considering the incentives a platform builds into its fabric we move from “how” we deliver community and media to “why” we are delivering the healthiest and most vibrant community our service can support. Getting that "why" right is the only way to ensure good intentions grow, rather than fade away.
Don't give us the thin veneer of a better algorithm, instead let us build better foundations for our online communities.
- Andy Toone
Back to My Writings