Human curation still matters
In a sea of AI-driven curation, the power of human recommendation is more important than ever. Particularly when it comes to connection and community.
Gilles Peterson ©Earth Agency
Rise of the machine
In the mid-2000s, Gilles Peterson's Worldwide broadcast every Sunday night from 11 pm to 1 am on BBC Radio 1. Each week, I would drive to the middle of nowhere (not a difficult feat when you live on the edge of the Peak District) and listen intently with a pad and pen in hand. No one else around, no distractions: just me, the music, and a page full of song titles hastily scribbled down so I could search my local record shop for them later.
That show – Gilles as the curator of a universe of music spanning obscure 60s Brazilian jazz to present-day broken beat and electronica, me as the willing student – had a profound impact on my musical education and has shaped many of my tastes to this day. It's still the first thing that comes to mind when I consider the power a good curator can wield.
At the same time as Gilles’s Sunday night broadcasts, algorithmic recommendation was gradually emerging online. So gradual that many of us - myself included - barely noticed it happening. It’s now over 20 years since Amazon started to implement algorithmic product recommendations on its site. Fast forward to the present, and the overwhelming majority of what you encounter online is dictated by an algorithm: from the posts presented to you on social media to the search results returned by Google. Not a human recommendation in sight. It’s not hard to see where the dead internet theory comes from.
Intentions matter
Many of these algorithms are fine-tuned hundreds, if not thousands of times annually to make their recommendations appear more human to us. The problem is that algorithms are programmed with much different intentions than your average human.
Take social media, for example. Your timeline on Instagram or Facebook is no longer linear: the algorithm determines which posts you see and which you don't. The intention here is not altruistic, but rather to hold your attention: keeping you engrossed in the app so that you continue scrolling and view more ads interspersed among the content. At its core, it's about commerce.
Laura Herman, a researcher at the Oxford Internet Institute summarised the potential cultural pitfalls of this in a recent conversation with New Scientist: “Instagram’s algorithm is deciding every second what millions, if not billions, or people around the world are seeing: what becomes their cultural reality. It’s just one single company deciding what people will see. The ripple effects of that are enormous.”
The value of connection
Human curation in contrast is often about emotion rather than data. It’s about the curator’s personal connection to the content they recommend. How it makes them feel. It's also about our relationship with the curator. The way they present and express themselves, how their interests align or diverge from our own, how they curate and share – all of these factors contribute to the experience. It's never as straightforward as simply the content itself.
Connection and community play a significant role. In the era of Discord, Substack, and social media comments, curation isn't merely about sharing; it's about fostering discourse. Worldwide, there are niche Discord channels where groups of passionate individuals spend hours discussing the finer points of their favourite films, art, or music. This is what genuine online community looks like, much of it centred around curation.
Human curation also encourages us to be challenged. Recommendations will often introduce us to unfamiliar territory: a new genre of film, an unexpected song, prompting us to consider something we wouldn't ordinarily be exposed to. Spotify in contrast isn’t trying to challenge your conceptions, it’s trying to keep you listening by recommending music that sounds as close as possible to what you already consume. While its marketing describes it as ‘discovery’ it effectively serves you with more of the same.
Human vs. AI: who wins?
Although the world seems to be awash in algorithms, human curation is actually experiencing somewhat of a renaissance. Just look at Substack: who would have predicted that an email subscription service would grow to 35 million users? Or 8 million Patreon users for that matter? Even on the micro level there are upstarts like Rovr, a platform attempting to reinvent radio while proudly declaring itself anti-algorithm. Clearly, many among us are seeking something more.
Humanity can even be found in places where you might expect the algorithm to reign supreme. In contrast to Spotify’s increasingly AI-driven approach (f**k an AI DJ by the way, in my humble opinion) Apple Music still emphasises the human element, according to its head of global editorial, Rachel Newman: “At Apple Music, human curation has always been the core to everything we do…Now more than ever, we know that investment in human curation will be key in making us the very best at connecting artists and audiences.”
Regardless of personal opinions, algorithmic curation is not going away anytime soon. Nor should it. Much of our online world is built upon its foundations, and although we may not view it with warmth and empathy, it undoubtedly has value in numerous aspects of our daily lives. It's scalable, functional, and reliable – precisely what it needs to be in many of the places it operates.
The key for us as human beings is not to think about whether we should use machine-driven curation, but rather when we should look for human input. If I require product recommendations when an item in my online grocery shop is out of stock, then AI is welcome to assist me. If I'm seeking a new piece of art to move or challenge me, I'll defer to a human like Gilles Peterson every time.