Creativity, Social Media & Selection Bias

— on May 14, 2019 by in editorial

Where do the ideas of artists come from? It’s an old question, and not surprisingly, different cultures have had different ideas about it. Our modern world loves to talk about the “AH HA!” moment where the lone genius synthesizes a great idea where none existed before. Surprisingly, little distinction is made between the genius of a physicist and that of an artist.

A photo from a black light rave soaked shoot we did a few years back.

A closer study of creativity, artistic or otherwise, shows a different story. Great ideas comes out of fertile environments, and often have a long slow gestation. Isaac Newton (of Newtonian Physics fame), Carl Jung (creator of analytic psychology), Albert Einstein (relativity and a lot of modern physics), Charles Darwin (evolution and natural selection) and many others arrived at their ideas through immersion in an environment full of others working on similar, or in many cases, dissimilar ideas.  Even when the creator believes it was an “AH HA!” moment, the facts show otherwise.

In fact, the modern idea of a “scene” (e.g. a music scene, a foody scene, etc) is that a lot of people are working on related things in close proximity. A city with a great music scene has a lot of musicians, all interacting with each other. Their ideas play off each other. To paraphrase from Matt Ridley’s TED talk, when you put a bunch of ideas in a space together, they have sex…

Social Media

In this light a system like Facebook looks like a perfect virtual location for creative scenes. It has the ability to connect us to many other people in a myriad of disciplines. In fact, Facebook, and other social media platforms should be the perfect fertile ground for creatives. I know I’ve certainly spent a lot of time on social media, and it’s helped me form many ideas.

But, I’ve been thinking a lot lately about what I want my social media presence to be, and it’s cost. It’s hardly news that social media is also time suck. While it is (potentially) fertile ground in terms of ideas, it might not be the most efficient fertile ground, and it certainly has costs in terms of attention and distraction.

Facebook’s (and other systems’) “algorithms” are designed to create engagement, which in short means showing you things you are most likely to interact with. How does it know what you are likely to interact with? It shows you things similar to things you’ve seen and interacted with before. Past determines future. That is a vast oversimplification, but a helpful one.


Imagine a coffee shop near a major research university. The room is full of intellectuals: professors and students alike. There are many conversations happening at once all over the room. The culture is such that anyone might overhear something and choose to bring it into their own conversation, or join the conversation it came from. This is the cafe culture associated with a number of enlightenment or particularly product periods in history.

At face value this sounds remarkably similar to Facebook but for the addition of The Algorithm.

Imagine instead that each person was in a sound proof room in the cafe rather than all sharing one large room. In the walls of each room there are speakers that can pipe in any of the other conversations going on.

Our creative, Sam, sits in his room and “browses” the conversations going on. He can’t hear all of them or even control which he hears. Instead a maître d′, let’s call him Al, is in charge of turning on and off the speakers from other rooms. The longer Sam sits, the more things Al will let him hear. Al’s job is to keep Sam sitting in the room drinking coffee so Al wants to make sure Sam is as engaged with other conversations as early as possible and stays engaged as long as possible.

A Bad Middle Man

Now, this might seem like a reasonable scheme, but it has a number of dangers as a creative space. The primary one from my point of view is that Al is likely to use Sam’s history, and in particular, what has kept him sitting there in the past, as a guide for what to let him hear in the future. Al, and Sam’s history are the gate keepers to what he is allowed to be exposed to.

The real danger of systems who’s metric for success is engagement (Al wants Sam to stay in the room drinking coffee) is that they don’t care HOW they keep Sam there, and the easiest ways to drive engagement are to filter conversations towards a specific types that grabs people. Since the only data Al has is the history, a recurring pattern is established.

Survival Bias

Beyond The Algorithm, there is another problem with social media platforms: bias and it’s effect on what we think of our own work.

There is a (much exaggerated) story from WWII where the British Navy was trying to improve the survivability of it’s combat aircraft. They did some basic analysis. Locations most often damaged on missions received extra armor. All very no nonsense and straight forward.

And, it had almost no effect on aircraft survival rates. The legend says that statistician Abraham Wald came in and told them to reverse where they put the armor, instead installing it on the locations least often damaged on returning aircraft. By analyzing only surviving aircraft, they determined which parts of the aircraft could survive a hit. The places least often seen damaged were the real weak points since they represented the lost planes.

For the real story, see this article. It isn’t as dramatic, but it boils down to the same lesson: be careful and conscious of biases in your data.

The Like Trap

What does this have to do with social media presence as an artist? Well, for me, it’s a cautionary tail. We need to understand what the data we use to make decisions means and why. Likes are the main culprit that I’m thinking of.

We all like a little (OK, a LOT of) affirmation and social media is great at providing that. In fact, one of the ways Al can make sure you come back is to get things you post some likes.

As a creator, the problem comes in when you start letting what your audience engages with (in particular “likes”) drive your future work. It’s easy to do even if (and inspite of) you consciously understand that a “like” is not a measure of how good an idea, such as a piece of art, is.

Likes are not a metric of “goodness” in creative works.

Likes are generated through a filtered process. Al is in the middle. He is going to show your work to people who liked previous work (after all, he wants them and you to stay engaged). Likes come with a HUGE bias as a measure of quality, especially if you are trying to break new ground.

Confusion and New Ideas

Social media, and in particular Al, work on a simple principal: keep people engaged by showing them things they are likely to interact with. What are people likely to interact with? Things that are familiar (whether they love or hate it). Familiar things, even that a person hates, require very little energy and are safe.

Low energy and safe you say? Lets do that!

What do a lot of new amazing ideas (and art) have in common? They confuse their audience. Viewers aren’t sure what to make of them. They don’t fit into the nice comfortable spaces in our brains. The idea that space was elastic and that an object going faster gained mass is weird if you only know newtonian physics. People are still fighting the idea of evolution today despite overwhelming evidence because it conflicts with their religion (i.e. what they were told by their parents).

It takes time for for new ideas that don’t fit in a viewers frame of reference to be accepted.

What is likely to happen to a new revolutionary piece of art on social media? Most likely it will get little or no engagement initially and Al will relegate it to the “didn’t generate engagement” closet. It won’t generate any likes.

As an artist, if I, even subconsciously, equate likes to success, I’m letting Al choose what art I make.

Art and Growth

I try hard not to set out to make photos people like (social media or otherwise). I hope they DO like them in the end, but that can’t be the goal up front. If getting likes were the goal, I’d post the same 4 archetypal photos over and over again. Instead, I set out to do something new each shoot. As a viewer of my photos, maybe that element isn’t obvious, but it’s there.

Only a few years back, I was pushing myself to make more and more interesting milky way and startrail photos. They were the most common type of photos I was making at that point. Any given new moon weekend I was probably out somewhere shooting. 

A creative milky way photo with an artistic zoom effect added in camera.

The reason I was focused that intensely on astrophotography wasn’t that I thought it was likely to get me the most likes. Instead, it was because I was trying to create new and interesting images through perfection of technique. Each shoot was a chance to push a little harder, or try something new.

Recently, I’ve slowed down. I rarely go out specifically to shoot astrophotography unless I’m somewhere special. Why? Frankly, I’ve explored that space in my art and moved on to other types of photos, and other elements in my photography. 

Conformism and Consistency

I’m not suggesting that there is anything wrong with producing consistent results in a specific genre. Some people photograph one type of subject their entire career, and thus share only that genre.

While I’m a genre hopper, the same reasoning that applies to me, applies to someone that shoots one type of thing all the time. Don’t let yourself fall into a trap of reproducing work that gets more likes for the sake of likes.

Let us take for example, a hypothetical infant photographer who only posts photos from sessions with newborns. This is their passion and why they pick up a camera. One day they subconsciously or consciously pick up on a trend… Photos of newborns with plush animals in them get more likes. 

So What?

Sharing work is a key part of many artists’, including my own, artistic process. Share early and share often. Share both successes and failures. For many working artists, sharing on social media seems like a primary driver of new business. This post might seem to conflict with the idea that sharing on Facebook and other social platforms can be beneficial, but it doesn’t.

Sharing on social media is fine if you are careful. I’ve personally left Facebook, and mostly stopped posting on Instagram because the danger of craving another hit of “likes” is to high for me. When I go to start a post the voice in my head is always asking “but will people like that one?” As a result I won’t let myself post just to have activity. I post only things I’m excited about, and not as a habit or part of my marketing.

Related Things to Watch and Read

I’ve been thinking a lot about work, focus and creativity recently. This blog grew out of that.

While not directly related, the book Deep Work by Cal Newport is mostly about focus and productivity. However, it touches on collaboration.

I’d be remiss if I didn’t mention both Steal Like an Artist and Show Your Work by Austin Kleon which include a great many ideas about collaboration and sharing.

TED has a bunch of good talks related to creativity with a strong basis in science. Here are a couple I’ve watched recently and like:

On creative process

On imposter syndrome and the pressure our culture puts on artists:

On being a multi-potentiate and it’s advantages to creativity:

On the unique way humanity recombines ideas that is more similar to genes than anything else: