“The aim of the painting is that the eye should find out what it likes.“

– Marion Milner

A recommendation is defined as ‘a suggestion or proposal as to the best course of action’. Whether it’s a stock-tip from a broker or more likely, online content served to you by an algorithm, recommendations help save you the time and energy that decisions often require.

Since time and energy to think is not something that most people have in abundance, recommendations are usually welcome. A point of view on what film to watch next or the quickest route home can come as a relief.

There’s no obligation to action a recommendation if you don’t want to. No one is forcing you to buy a book ‘selected for you’ or to cook the ‘recipe of the day.’ But by definition, they infer an action to take. For example, after a well known author recommends a novel, you might decide to sample or buy the book. You might decide to watch a Western film after Netflix recommends the genre to you. And after a spirited recommendation by a sommelier, you might drink the wine. These actions, whether online or offline, big or small, subtle or obvious, reveal the other side to a recommendation.

A few decades ago, a friend might of suggested a recipe to cook simply because they liked it, or because they thought you might like it too. Now, a featured recipe considers dozens of additional factors and data points like global trends, your cooking history or context like day of the week.

It’s simple. If a recommendation understands and meets your needs, you’re likely to choose it again. If you like every song Spotify thinks that you’ll like, you’ll keep listening, right? If you’re going to see ads on your YouTube videos, wouldn’t you rather see “ads about the things you care about?” The better they get, the more you rely on them.

With the help of AI, Mark Zuckerberg says that soon Meta will be so good at ad recommendations that “any business… can just come to us, not have to produce any content, not have to know anything about their customers… and pay (us) for as many business outcomes as (they) can achieve.”¹

---

The serendipity of a new song or the salivating promise of hyper effective ad targeting are non-trivial benefits. But one could also benefit from spending equal or more time exploring alternatives, like figuring out what you actually want.

It’s an important, but difficult question to answer. What makes you happy? What do you dream about? What do you obsess over, and why? For a week last winter, for reasons unknown, I got interested in the history of Samurais. Ironically, by the time I was getting served film recommendations, I couldn’t have cared less about them.

Threads of interest, curiosity and delight are unlikely to be directly leveraged by algorithms and but are frustratingly difficult to notice for yourself. In the 1930’s, British writer and psychoanalyst Marion Milner wrote about her efforts on this problem, which sent her to almost Carl Jung depths of the unconscious. But the journey was worth it.

I had also learnt how to know what I wanted; to know that this is not a simple matter of momentary decision, but that it needs a rigorous watching and fierce discipline, if the clamouring conflict of likes is to be welded into a single desire. It had taught me that my day-to-day personal “wants” were really the expression of deep underlying needs, though often the distorted expression because of the confusions of blind thinking. I had learnt that if I kept my thoughts still enough and looked beneath them, then I might sometimes know what was the real need, feel it like a child leaping in the womb, though so remotely that I might easily miss it when over-busy with purposes. Really, then, I had found that there was an intuitive sense of how to live. For I had been forced to the conclusion that there was more in the mind than just reason and blind thinking, if only you knew how to look for it; the unconscious part of my mind seemed to be definitely something more than a storehouse for the confusions and shames I dared not face.²

The sort of introspection Marion conducted is absolutely not for everyone. It’s certainly easier to follow recommendations from other people. But a recommendation will always be what someone else thinks is best. What do you think?

1. An Interview with Meta CEO Mark Zuckerberg About AI and the Evolution of Social Media ↩︎
2. Marion Milner – Goodreads ↩︎