Last week, I got excited reading this post from .
Temperature Zero for Culture: Why Everything Is Starting to Look the Same
Perhaps I am late to the party, but since it took even the NYT’s 3 months to get to it I don’t feel that bad.
In her post, Leek explores how algorithms progressively concentrate attention on what is already familiar and successful. Recommendation systems can narrow our cultural choices by repeatedly rewarding what is most predictable. Tattoo that on the back of your hand, will you? The more algorithms optimize for predictable engagement, the more they reinforce existing preferences, reducing the opportunities for unfamiliar work to develop an audience. Do you need a clearer explanation for why you struggle to find what you want on the major streaming platforms?
OUCH! Right? I bet you instinctively suspected this already. It is the algo. We all feel it. I regularly hear people suggest as much. But if you like data to back up those feelings you’ve been having, Lauren’s got some for you.
Using a dataset of nearly a million publicly posted film ratings, Leek found that just 1% of the films accounted for nearly 40% of the recorded activity. Chew on that. We have access to more movies than ever before, but our attention is concentrated on a remarkably small fraction of them. Despite great access, we have less discovery. Now, those ratings don’t establish what everyone actually watched, let alone prove that streaming algorithms caused the concentration. But the pattern is important, particularly when considered alongside her simulations of how recommendation systems reinforce existing popularity. We get stuck in the rut of the same old same old.
Remember that dream we had earlier on, the one about how streaming could be something different than what we currently have? We were going to have access to practically everything. New work and old. Films from everywhere. A magnificent smorgasbord of delights. And the discovery? Oh my, it was going to be endless delights. Instead, we just encounter variations on whatever everyone else is getting fed. I don’t know about you, but I want a system that expands my taste and introduces me to things I never imagined existed. I don’t want my viewing history to become a prison of my own predictable preferences… which is what has happened to us all.
But What If 20% Could Change Everything?
And here’s where Leek’s research really got me excited. Her simulation of algorithmic recommendations suggests that preserving a broad range of cultural choices may require platforms to devote substantially more attention to exploration than they currently do.
In her model, forcing an algorithm to explore unfamiliar options with only 2–5% of its recommendations accomplished very little. The available range of choices continued to narrow. But at approximately 20% exploration, that narrowing began to reverse. The algorithm had to devote roughly a fifth of its recommendations to unfamiliar choices, rather than just pushing more of what its data told it people were already likely to watch.
That’s a simulation, of course, and we can’t assume 20% is a universal threshold for every streaming platform across each and every cultural strand. But it raises a good question. If algorithms need a substantial commitment to exploration to preserve a broad range of choices, shouldn’t the companies that increasingly control our access to cinema be required to make that commitment? I bet you can guess where I stand.
And if we’re going to require them to devote more attention to unfamiliar work, shouldn’t we also require them to support the independent suppliers who make so much of that work possible?
This becomes increasingly important in the current discussions around a new version of the Financial Interest and Syndication rules (FinSyn) for the global streaming platforms (GSPs). Those earlier regulations addressed the concentration of ownership and control in television. Today’s platforms have another tremendous power: they can influence what gets discovered in the first place. They are already have been colonizing minds, tastes, and preferences worldwide for over a decade.
And let’s also not ignore how availability is different from attention. A platform is able to stack 20% of its library with “independent” cinema while also ensuring that virtually nobody encounters it. A quota based exclusively on catalog size would accomplish almost nothing. A requirement affecting what gets recommended, promoted, discovered, and ultimately watched could begin to change the economics of independent film. We need to consider the big picture here. And hopefully the big wigs already lobbying for a new FinSyn have this in mind already… but who knows?
If we want authored cinema to remain culturally adventurous and economically sustainable, merely making “independent” films available will not be enough. We need to ensure the type of films we love have a fair opportunity to be discovered, valued, and financially rewarded.
I reached the end of Leek’s post pondering whether we could establish a 20% minimum for genuinely independent, third-party supply on major platforms, while also ensuring that audiences regularly encountered a comparable percentage of films outside the most familiar and predictable choices. Could that help us build a more sustainable cinema ecosystem? Wanna bet again where I stand?
Leek’s research doesn’t establish that 20% independent supply or licensing expenditure is the correct minimum. Those are propositions I’m adding on to her finding. But if we accept the possibility that cultural breadth requires a meaningful commitment to discovery, shouldn’t we also examine the economic conditions necessary to maintain it?
After all, what are we actually building toward? I want a cinema ecosystem in which a wide range of creators can continue producing work, audiences can discover it, and independent businesses reinvest in future films. EZPZ. That requires something more than oh so many of us getting a film made once. It means creating the conditions that allow the people who make and distribute films to build sustainable creative lives. And that should be the priority of all of our Film Support Orgs (FSOs).
And I think that’s going to require some regulation, both stateside and globally. Let’s get into those now, shall we? Maybe we can put a few things on some people’s #ToDo Lists.
Who Said the Platforms Get to Make All the Rules?
Read more
