👮 paul blart, mall cop

When data deadens culture

May 15

Optimist: The glass is ½ full. Pessimist: The glass is ½ empty. Excel: The glass is January 2nd.

it’s so fucked up that if you get a burger from anywhere besides Five Guys you have literally no way of knowing how many guys were involved in that process

Columbia business school post-doc Adam Mastroianni recently outlined how Pop Culture Has Become an Oligopoly; pointing to invasion, consolidation, and innovation as forces that have made culture familiar and repetitive - boring.

Adam Mastroianni @a_m_mastroianni

Chart-topping original movies have gone extinct. People have a lot of explanations for this, but they're all incomplete because they don't realize the same thing is happening everywhere. An oligopoly has conquered all of popular culture.

His observations felt counterintuitive, because consumers of film, television, music and books have more “choice” than ever before. In theory, the diversity of streaming platforms and the ability for artists and authors to self-publish should lead to more competition that ultimately rewards cutting-edge creations.

Instead, we keep getting comfort blankets.

Gesturing at the potemkin of choice ignores two critical facets of how a lack of competition is constraining culture: the first is the critical role of gatekeeping intermediaries in distributing cultural content, and the second is how consumptive data may be used by incumbent platforms to de-risk audiovisual decision-making that optimises for eyeballs. 👀

1. The role of gatekeeping intermediaries in distributing cultural content

For instance, Spotify is essentially vapourware for the three major record companies - which take in about 80% of the revenue from streams on the platform.

steve albini @electricalWSOP

1/ There's an important thread of continuity over time about the exploitation of bands by record labels that deserves a closer look, re the current Spotify debate.

occupy alternative rock @Eve6

our stupid band gets close to a million monthly streams on spotify. spotify pays out .003 cents per stream. 100% of that goes to our former label sony who is a part owner of spotify. this is why i’m mad

Last fall, radio DJ Patchen Mortimer reminded the internet how a 1996 change to the US Telecommunications Act changed ownership rules that facilitated consolidation across radio station ownership, culminating in one dude in Texas reshaping the airwaves. Intermediaries matter.

Patchen Mortimer @patchdj

But something else happens in 1996: the Telecommunications Act, which is good for a lot of things but loosens the restrictions on who can own how many stations where. 8/

Movie distribution rights are tightly guarded, and firms like Disney are so focussed on box office release numbers that they won't even let movie ushers see their films for free.

Margaret "Molly" Rasberry (she/they) #BLM 🏳️‍🌈 @RasberryRazz

One of the perks of working at the cinema are free movies except for Disney which it threatens to remove its movies from the theater if an employee sees one of their films during opening month for free. We had to wait a month or buy a ticket to see a Disney film. Pure greed.

How Cineplex screws over independent cinemas was recently summarised by Kingston-based independent cinema The Screening Room.

Anna Soper @Flora150Project

I was today years old when I learned that my local independent cinema (the only downtown cinema in #YGK) cannot screen first-run films until Cineplex has wrung all the $$$ it can from exclusive theatrical runs of films truly meant for cinemas like The Screening Room.

The Screening Room @screening_room_

They are able to use their dominant position in Canada to demand that distributors give them *first choice of every single movie* and *exclusive theatrical runs*, so competing cinemas like us need to wait until Cineplex is "done with it" before we can open a film.

These instances remind us that while great art may be made, there are huge hurdles to access markets through distributive platforms. Which brings me to the second point: the distributive platforms that have this gatekeeping power increasingly produce content that directly competes with artists that lack access to the data that may be used to make these decisions (so many words in 1 sentence, I know).

2. How consumptive data may be used to de-risk cinematic decision-making that optimises for eyeballs.

Amazon is emblematic of a theoretical ability to integrate insight(s) from across its various verticals that can inform the form and timing of cultural production through its in-house Prime Video or newly-acquired MGM Studios. In this context, the consumer is a critical data input and source of valuable information. Due to data passively volunteered by the consumer, Amazon knows things like:

What books people are buying (and pre-ordering) through the Marketplace and Kindle platforms,
How fast they are reading books, if they are reading the entire book, and what passages people are ‘highlighting,’
Whether people like a book and what they like about it (Amazon owns Goodreads) - this includes whether they are consuming it as an audiobook (Amazon owns Audible),
What actors and directors people appreciate (Amazon owns IMDb),
What translations are popular (Amazon Crossing).

The quixotic aspiration of using data to determine artistic decisions was recently parodied in Jennifer Egan’s “The Candy House,” where a character is hired to turn stock elements of movies and television into algebra and summarised by Chris Jones, author of “The Eye Test: A Case for Human Creativity in the Age of Analytics.”

Relativity Media was backed by ~$2B in liquid assets via Elliott Management, a hedge fund. Film financier Kavanaugh promised his investors he would make only hits using analytics. This was during a time that movies were becoming increasingly expensive to make. A more ‘demand’ driven model would, in theory, allow the production company to ‘safely’ create media that had a higher probability of being well-received by audiences (via the box office).

“Data” is how the world got “Paul Blart, Mall Cop” - a film poorly received by critics but earning $183M USD at the box office. It was “data-driven,” and producers anticipated that it would have mass appeal, and they were right. Yay?

Today, Amazon is emblematic of the hypothesis that motivated Kavanaugh and may be redefining how art gets made in a digital economy. This means that we should:

consider the anti-competitive implications of horizontal data integration and ;
care more about whether and how the use of data-driven insights can fundamentally de-risk the commissioning and/or creation of cultural content decisions that are informed by rich derived information that independent directors and producers cannot proxy.

Netflix’s House of Cards was an early example of this approach: the firm knew that the British version of House of Cards performed well and - at the time - media with Kevin Spacey or directed by David Fincher also performed well.

Netflix’s recent foreshadowing that it will seek to curb password sharing has been framed as a mechanism to recoup lost revenue. Less sharing also means more accurate algorithmic information that can be used to tailor movie trailers and recommend shows.

Firms seem increasingly interested in automating or predicting culture. As an example. Canadian unicorn Wattpad’s recent Grand Plan references the data and analytics generated from stories as a tool to turn “stories into hit movies, TV shows, print books, and other media.”

The world doesn’t need another Paul Blart film, but the ‘Paul Blart approach’ is a reminder that we might need more ‘mall cops’ in competition that can address manipulative gatekeeping and exploitative data usage. 🚓

🏢 regs

In the US, the FTC recently invoked “algorithmic destruction” as a way (remedy) to penalise companies for unfair and deceptive data practices. The basic premise is that a firm will have to destroy algorithms of AI models that are built from information that was improperly obtained.

This intervention is a reminder of the overlapping intersection(s) between privacy and competition policies and whether data is a ‘secret’ weapon or stolen asset.

Are consumers of content consenting to their viewing history (+ psychotropic analysis, mood analysis, emotional analysis), reading information and music streams informing cultural production decisions that enrich platforms at the expense of creators? 🤔

PS. Parliament took a look at remuneration models for artists and creative industries in 2019. The briefs are fascinating and largely focus on copyright. Below a quote from Cory Doctorow’s brief:

Heritage must work with Industry to end 40 years of antitrust enforcement based on the "consumer harm" standard of the Chicago School. This malpractice has led to massive concentration in the entertainment sector – as well as most other sectors – meaning that creators are sellers in a buyer's market, with precious few bidders for their creative output. It is elementary economics that sellers in a supply chain with few buyers clear less than they would if they had a larger pool of buyers to play off against one another. The long-run practice of allowing mergers and acquisitions in the media sector has produced so much concentration, both horizontal and vertical, that creators who succeed are still doomed to take home a smaller and smaller share of the revenues generated by their creative work.

regs to riches