‘Cuties’, Netflix, and Computation

In September, Netflix added the film ‘Cuties’ and well…there was a scandal. Soon #CancelNetflix was trending on Twitter.

Amongst all this lots of folks were asking my opinion about the Cuties poster. I actually have never had so many people ask me to write an essay before.

A lot of this was spun out from some of my previous work and from me telling some folks that I would love to work at Netflix after I graduate in June (if anyone from Netflix is reading this please please hire me when I graduate).

The spark of this fire was this poster, very different than the French poster (the movie is a French film) on the left.

Left: the original French poster, Right: The Netflix promotional image. Source: Deadline.com

When I was being asked about this it came down to these 5 questions:

1. “Nina…Nina…Nina have you seen this?!”

2. “Nina, is this scene from the poster actually in the movie?”

3. “Nina did they photoshop the actors or edit their bodies?”

4. “Nina…Nina…WHY IS THIS?!”

5. “How do we…not do this again?”

This post seeks to offer some answers (not all the answers) and context for these questions. And to explore how some decisions around this poster came to be. And to understand the visual ecosystem this exists in. And to offer some framing and suggestions to do better, along with some technical analysis of some of the materials.

1. “Nina…Nina…Nina have you seen this?!”

If by “this” you mean the movie — yes I have watched the movie and seen the poster and controversy around it. So…you can stop sending me DMs about it.

For those unfamiliar, here is a summary of the film, from this article I read from Caira Conner that I think people should be reading before they take Ted Cruz’s opinion too seriously.

It follows an 11-year-old girl named Amy who, like Doucouré, is the child of Senegalese immigrants. As she struggles to accept her family’s cultural and religious traditions, Amy befriends a group of girls who are part of a dance crew called the Cuties. Hot pants, hair dye, and twerking ensue. The film also grapples with class, race, and the role social media plays in shaping young women’s self-image, but that’s not what pulled it into a cultural firestorm. (Caira Conner — Watching the Outrage Over Cuties as a Survivor of Pedophilia)

I actually liked the film. I could tell how much work and research went into it. The acting was great and I thought it was well written.

Cuties thoughtfully portrays Amy navigating the exhilaration and cluelessness that can emerge between true childhood and early puberty — especially in an era when young people spend so much time on platforms such as TikTok and YouTube, where sexualized content abounds. While researching the movie, Doucouré spent a year and half interviewing more than 100 Parisian girls about their experiences with social media so that she could better understand the forces that shape young women today. (Caira Conner — Watching the Outrage Over Cuties as a Survivor of Pedophilia)

You could tell the director, who actually faced horrible threats from the controversy, spent so much time researching and telling such an important story. It felt complicated and nuanced instead of those overly fluffy coming-of-age films that always have a love interest. I think the author I quoted above sums it up best here:

It’s a movie about girls not wanting to be little girls anymore. About girls spying on classmates in the bathroom, girls who want to wear tight pants and hump the dance floor, without understanding why no well-intentioned adult wants to look at that, and without understanding the danger of the adults who do. (Caira Conner — Watching the Outrage Over Cuties as a Survivor of Pedophilia)

So yes, if I’m honest, I think that the film is a ridiculously important for pre-teen girls of this generation. A generation that is growing up very differently and in unprecedented times in all the ways.

It also shows and handles race and cultural clashes within the friend group and families. I certainly don’t remember these types of films from when I was a teen and I don’t think there are many now that cover so many important things as this film does. I think it immerses people in this difficult and strange age of girlhood instead of condescending it like some films do.

And yes, there are scenes that are…difficult. And I understand why people are so concerned, upset and even for many triggered by them. Again, I highly encourage anyone to read Caira’s piece, because I think it tells a side of the story I haven’t seen enough, but it’s going to be a different experience for everyone and that is ok. Also, it’s not really what I am here to comment on.

2. “Nina, is this scene from the poster actually in the movie?”

Not directly. The costumes are and the 4 girls do have a dance number in them. My assumption is that this was either from a cut scene with some photo edits or a promotional photo that was taken. Or from 4 separate stills of the girls that were then cropped and put on a background.

I believe that these were chosen because it’s definitely the most “dance troupe” vibe costumes that the show includes, which was probably the original angle on the Netflix side. And if I think about American dance or cheer movies — the bright outfits, ensemble in a strong end stance that represents the number — that is textbook visual for this type of media.

Posters from “Bring it On” movies, popular American cheer movies. You can probably see the visual parallels where the Cuties poster draws from these types of media.

The issue here is that this scene in the movie is by far one of the most provocative, and thus so are the stills. It also isn’t the core of the movie — the movie really isn’t an American style cheer or dance competition movie — it’s a coming of age film about Amy.

3. “Nina did they photoshop the actors or edit their bodies?”

I mean. Yes. I don’t know what to tell most people but most things you see on the internet are edited. There is a whole visual vocabulary in exactly this.

You can clearly see this from just the photo editing and how it has affected the skin tones of the actors.

Left: The average skin tone of the actors in the French movie, Right: The average skin tone from the Netflix promotional poster. Please note: These are rough averages and are not meant to be representative.

“But Nina — in the scene on the right they are under lights!”

You are right! However, skin reflects light a lot differently than other materials. Even super bright spotlights! It takes a lot of lumens though to go all the way from the 400 to 300 range of Fenty products.

For the girl on the very right (actress Esther Gohourou), this is a comparison of her average skin color from the actual shot in the film next to her skin color from the Netflix promotional poster.

Average skin color from the movie scene and average skin color from the poster of the scene for the right side actress, Esther Gohourou.

The poster on the left (the French release) is more grey toned and the girls are under some tree covering, adding shadows and skewing the average.

Average skin color from the movie scene, the poster of the scene, and a direct sunlight scene for the right side actress, Esther Gohourou.

So I added in a swatch from a scene where they are in direct sun to see the difference between the dance scene, poster, and a direct sunlight scene.

Basically, whatever filter and/or edits were applied definitely warmed and lightened the skin.

They also did some major retouching, etc which I sort of expected. If you were ever an 11 year old girl with puberty skin and cake foundation, you know damn well that is not what skin looks like. And no, the makeup isn’t as snatched on film as it is on the poster.

I actually don’t know how much the bodies were photoshopped. I believe they were retouched and the color space of the photo was changed with lots of different edits, bringing different emphasis to different parts of the image.

Also please note, these swatches are weighted and normalized averages from the pixels of the skin of the subject — they are meant to show trends but not fine grain analysis. Analyzing skin color in photo and video is actually a super difficult and nontrivial task and there are defintely better ways to go about it.

But also…why do people want to know if the girls were photoshopped?

Last year I released a few things about how digital media is effecting how we see ourselves and what we present ourselves with in our social media spheres.

I think in general this desire to know if something is or is not photoshopped comes from an increasing digital distrust and instability of our perceptions of ourselves. Overall, after thinking about how people look for a while I have realized that a lot of people are losing grasp as to what real skin and real bodies look like in digital space.

We are so visually saturated but this visual saturation comes with uncertainty. It’s like swimming in fog.

We also exist in a world of hyper commodification of…everything. And I think there is also a very important relationship that needs to be interrogated between the body and how it is displayed in digital media. And in particular, how this intersects with different identities. After all, this very visual communication has helped enable segregationist ideals and active discrimination against people of color, but that is a different essay.

And these girls exist in bodies awaiting a different politicalization than bodies like mine and others writing and arguing about what they do with theirs.

That said, I think it’s also important to understand that this is a French film. The way that race and class and culture are seen are different than in the US. And perhaps it is with this reminder throughout that we can also understand why these visuals and marketing and responses were so different as well — most people reading this are probably American and also have American bias like myself and a lot of the staff at Netflix. And I can’t comment towards it all or sort it for you, dear reader.

I exist in a white, American body. I have an invisible illness, and while my medical conditions and rights are politicized (especially in a pandemic)…I don’t often find the presentation of my body politicized. I am a cisgender woman who is a size 12 and everything about my presentation aligns with an acceptable presentation of white femininity in the United States. I am immensely privileged in this. I am also in my twenties and there are very few photos of me at age 11.

And girls in this age range face a different and faster sexualization than many in my generation. I say different because this is not new. The sexualization and exploitation of underage girls is not new. It’s not. The New York Times showed last year that it is a digital breaking point, so acting like this movie is a catalyst isn’t entirely a basis.

However, we didn’t have smart phones our whole childhood and we didn’t have hyper visual spaces like TikTok and Instagram where we were mingling in the oceans of all different ages. So yes, things are different from our ages.

4. “Nina…Nina…WHY IS THIS?!”

When I was about 11 there was one distinct *look* I wanted: an icy pink lip and straight only sorta wavy blonde hair like Carrie Underwood. I gathered all of this from a single cover on a magazine that I saw at the store.

I screamed and badgered my mother about this single image for months. It visually stuck in my brain.

That is what magazine covers, billboard ads, Instagram posts, and Netflix promotional materials are trying to do. They want images to stick like a stubborn price tag to a cheap wine glass you bought at the Dollar Tree because your life was falling apart.

I have always been insanely impressed by Netflix’s marketing and promotional media. I especially love their spooky season twitter accounts (bless whoever coded and helps run the Haunting twitter account) and how they market the Haunting anthology series visually.

A few months ago I found this article talking about some of the science that goes behind this — there really is a lot of amazing things that go into these decisions. There is also an insane amount of personalization and localization that go into these promotional still decisions all across the Netflix app.

And they need to! After all, they have mere seconds to secure a viewer.

Promotional materials on Netflix, from what I can tell, are very very filtered. This also makes sense.

Our eyes aren’t as used to looking at stills from films. These stills often have to be filtered and edited to stand out as a still promotional image. Especially if they are being used for hyper filtered and saturated spaces like social media or even within Netflix’s app, often looked at after one of these types of feeds.

Filtering is actually something that we can detect with a variety of computational approaches. I mentioned a project I did regarding filtering and photoshop in social media in Ethan Zuckerman’s class this past spring. We found, unsurprisingly, that most pictures of the influencers we looked at were filtered or edited in some way.

Some previous work I have done on filtering of images, as per Ethan Zuckerman’s class “Fixing Social Media”, viewable here: https://ninalutz.github.io/HistoryFilters/

So we are saturated in this sea of filters and color editing and retouching. But how do these photos get so filtered that the skin tones are so skewed? And how did this visual communication get so far off?

The first question is a much easier answer:

Photography and film making have entirely different rules and colorspaces.

I talked about this briefly above, but in general, photography and visual communication we have built around it, such as these promotional stills and your entire Instagram feed, is still reckoning with a racist and colorist past.

For many, this was highlighted this past summer in June and July with #VogueChallenge and with criticism regarding a cover shoot for Simone Biles on Vogue’s cover this summer.

This isn’t just about representation, though. This is about recognition and dignifying beauty of all bodies through visual media.

Fortunately, there are a lot of ways to think about this from the computational side and to bring this into a variety of systems. I am focusing on this one example, but there are people like my friend Dr. Katlyn Turner, who focus on frameworks and methods for designing and building antiracist practices and technologies across fields and identities.

For this instance, we know that the brain is attracted to bright, saturated colors and hues. And there are ways to pull these out and warm up whole scenes, especially in films that are darker lit or in shadow or other visual obstacles, without drowning out or losing a variety of skin tones.

In Computer Science we call one of these processes “Color Quantization” where we can determine mathematical distributions of colors from the pixels in an image and see them in different color spaces, such as RGB, HSV, HSL, etc. I actually spent the better part of last Spring term working towards a tool regarding developing more representative color palettes of photos with people in particularly (hint: skin is often best exposed in HSV or HSL spaces, so those should be referenced in these edits).

This is just one method, but it’s my personal favorite because it can give us a mathematical sense for understanding color trends in an image and where to filter and what colors we might lose in different filters. You just graph the axis of your color space (Hue, Saturation, Light or Red, Green, Blue, etc, etc) as geometric axes and plot pixels at their respective coordinates.

It allows you to see how the different spaces of your image are changing as you edit, and you can more fine tune these edits towards various priorities while also maximizing for those loci of bright colors to catch the eye of users.

Different color space layouts for the Netflix Cuties poster with palettes of the centroids of the most dominant color clusters (K-means clustering).

But it’s not just editing that went through here, since the poster itself didn’t represent the film in many ways and was a composite of some kind. This is the answer I don’t completely have.

I have some theories I mentioned above — such as the bias to fitting this visually striking and colorful part of the film into the visual mantra of cheer and dance team posters like “Bring it On”, but I think there are empathetic things to consider to ensure these instances happen less in the future.

5. How do we…not do this again?

I mentioned in the section above the magazine cover I saw when I was a child.

A little over a decade later, at 23, I read “The Ones Who Walk Away from Omelas” by Ursula K. Le Guin in an office I no longer occupy.

I read it as part of Dan Novy’s “Science Fiction Inspired Prototyping” class I took my first semester of graduate school. This was the same semester that my first advisor was fired for sexual harassment and my department fell apart under multiple scandals, including Jeffrey Epstein’s investments in the lab.

I cried reading the story in my office. Later in class, Dan choked up a bit talking about how it gives both the most simplistic and also poignant allegory of an innocent child suffering so that the people of this town could be happy and they were all made aware of it in a ceremony of sorts and yet…very very few resisted it. Very few walked away. As we sat in a room in a class in a department that had received money from a monster.

And that is the thing about stories and the worlds built within them. They feel both so separate and so connected to the tragedy we have made in our world.

This is why so many stories stand the test of time. We have built so much human society around stories.

And to ignore the power that visual media and stories have would be a disservice and denial to how much of society is organized and how we can communicate change. Whole cultural revolutions and political campaigns have been spurned from visual media, often film and photography.

We can all probably agree this entire thing wasn’t great. I am sure folks at Netflix had many a late night. I am sure a lot of people feel really horrible. And the director has received literal death threats, especially since women of color tend to be at much higher risk of online harassment.

So we…don’t want to do this again. At least not this way.

We want more diverse stories. We want challenging narratives by amazing creatives that are not always highlighted. We want curation of lots of stories. But…we don’t want this.

We do not want the weaponization of these stories because of errors of visual communication when new visual vocabularies are needed.

Most of the methods I used in this article were just some examples to expose and try to parametrize some of the systems and trends I saw within this instance. And to try to contextualize them.

But there is so much more here that needs to be considered in the treatment of these stories, especially when they are ones that are not always seen in mainstream American media.

And, if I’m honest, color theory and photography and technology aside, one of the best methods that may have avoided this could have been some kind of semantic mood board of sorts. Exactly the types of “aesthetic” collections that Gen Z is posting on TikTok and that millennials were posting on Tumblr.

Example of a Mood Board in instagram styling. Source: https://www.pinterest.com/pin/341992165447876661/

And while it feels silly to think about a “mood board” solution, the “mood board” solution is a manifestation of human visual grouping and logic, a huge cognitive skill that machines are still learning. And something that we can all intuitively understand.

Context comes from both history and also the context of promotional materials, marketing, and visual communication. But also the context of the current and future audiences, and a visual language that we all know.

6. Conclusion?

I believe in going to the source. And during the controversy that transpired, the director said this to a publication:

“I’m hoping that these people will watch the movie now that it’s out. I’m eager to see their reaction when they realize that we’re both on the same side of this fight against young children’s hypersexualization.” — Maïmouna Doucouré (https://zora.medium.com/the-director-in-the-middle-of-the-cancelnetflix-backlash-speaks-out-90b58f5afc64)

This summer my research group, which I joined in April, launched a participatory art project. I had never been part of an art launch, especially within the time crunch we were in, the scale, and being so new to the team.

During this time, I wrote an essay called “A nonanswer to “Art in these times”. I talked about struggling with the value of art from privilege and during a global pandemic while also staring down my own burnout.

What I learned at one point while helping my advisor edit promotional materials is that there is a lot more than people see that goes into the promotion of creative things. It’s a whole army of strategy.

So, I don’t have a concrete answer as to why this all happened. And I can’t say it will never happen again. I have some frameworks and technical suggestions and some insights into some of this issue, but I don’t have all the cards.

Netflix is a major curator. And at the end of the day, their biggest goal is to make you look. And bright colors, faces, compositions that draw in the eye — those are all going to make you look.

But what we are perhaps starting to realize that not every look is the same.

The old adage of “if it bleeds it leads” and “sex sells” might not be the same anymore. And even if we aren’t actively seeking these grammars, they might be ingrained into us with other biases.

And it might be time to try new things and I think there are frameworks and methodologies, some of which I used and referenced here, that can lead to better representation across all types of mediums.

Because like it or not, visual media and art and entertainment have power. Real power and influence over our minds and the world. And I think there are so many amazing people and stories opening new doors towards using art and the visual fog we find ourselves in to challenge and better us.

And just last year I was writing 9 dimensional systems of equations to describe how light behaved in real fog. So now, I only hope to spend my next chapter pursuing this metaphorical fog.

Nina M. Lutz is a student at the MIT Media Lab and thinks about human quantification, aesthetics, and data-driven visual communication. Previously, she studied Computer Science and Design at MIT for her undergraduate. She is currently working on participatory digital artworks for her MS thesis.

Please hire her after June 2021. She is so hireable and will do code and math and writing and art for you for compensation and health insurance.

She is currently being held together by TV and charcuterie and her friend Kelsey who is willing to drive her around in circles when she gets sad about the world. Her spirit animal is a 13 year old white dog who lives in the suburbs. Not because she has one but because it’s the life she wants.

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store