What the Control Room Knows
There is a moment in live television that most people will never see, because it happens a few seconds before the thing they’re watching exists.
A director calls a camera. An audio engineer opens a channel. A producer has made decisions about which version of reality will go to air, and why, and how fast. The audience receives the result of all this—the polished, continuous, apparently effortless thing on their screen—with no particular reason to wonder about what preceded it. Why would they? The broadcast doesn’t look like a series of choices. It looks like the news.
I spent twelve years teaching students how to work in those rooms. Now I study what comes out of them.
The formal version of my story goes something like this: I am a new assistant professor at the RTA School of Media (The Creative School) at Toronto Metropolitan University, one of the best media programs on this continent, in a city that has more working journalists than anywhere in Canada. I research misinformation, artificial intelligence in journalism, and a concept I’ve been developing called democratic legibility. I also teach production—multi-camera, audio, documentary, live sports, lighting—because I spent more than a decade doing it professionally while teaching it, and because I believe, with increasing conviction, that production knowledge is not a technical specialty.
It is a way of seeing.
The less formal version is that I keep asking a question I can’t quite shake, and this work is my attempt to work through it in public.
The question is this: can citizens actually read democracy anymore? Not in the ceremonial sense—not whether they turn out to vote or follow the news in a general way. I mean something more specific and, I think, more alarming. In a media environment defined by algorithmic feeds, AI-generated content, hollowed-out newsrooms, and platforms that are economically incentivized to privilege engagement over comprehension, do ordinary people have any realistic chance of understanding not just what they’re being told, but why the information they receive looks the way it does—who made it, under what pressures, with what tools, toward what ends?
The answer, I suspect, is increasingly no. And I think we’ve been measuring the wrong things trying to understand why.
The study of misinformation has produced an enormous body of rigorous, important work in the past decade. Researchers have mapped how false claims spread across networks, tracked the relationship between partisan identity and belief in conspiracy theories, and documented the asymmetry between the speed of misinformation and the speed of correction. Some of this research is genuinely elegant. Most of it treats the production of information as a black box—a thing that happens offscreen, before the data begins.
This is the gap I want to fill.
What happens inside the black box is not incidental to why information misleads. It is often the whole story. A newsroom running at half its former staff, under a ninety-second deadline, using a summarization tool it doesn’t fully understand, producing a headline optimized not for accuracy but for the click that keeps the lights on—none of that shows up in a content analysis. None of it registers in a misinformation database. And yet all of it shapes what a citizen receives when she opens her phone and tries to figure out what is happening in the world.
I know what happens in those rooms because I taught the folks that staff them across this country. I know what gets cut for time, what gets simplified for a general audience, what gets framed one way rather than another because the broadcast only allows for a 90 second package to air (plus the billboard/announcer’s intro) and there is no time to be more precise. I know that the way information looks is not neutral. It is the accumulated residue of a thousand small decisions, most of them made under constraint, by people who are doing their best in a system that makes their best harder to achieve every year.
Understanding those decisions—that system—is, I believe, essential to understanding why democratic legibility is in trouble. Not because journalists are bad at their jobs. Because the conditions under which journalism gets made have changed in ways that the research hasn’t fully caught up to.
Democratic legibility, as I am developing the concept, is the capacity of citizens to read, interpret, and act upon the information systems that govern democratic life. The phrase inverts a famous formulation by the political scientist James Scott, who argued that states can only govern what they can make legible—what they can see, measure, and categorize. Scott’s legibility flows downward, from institution to citizen: the state looks at you and makes you readable to itself.
What I’m interested in is the reverse. Can citizens look back? Can they read the institutions, the platforms, the newsrooms, the algorithms that are constantly producing the information environment they inhabit? And if not—if the system has become so complex, so fast, and so deliberately opaque that ordinary people have no real way of knowing what they’re being given or why—what does that mean for democratic self-governance?
These are not rhetorical questions. They are, I think, empirical ones. And the fact that no one has yet developed a reliable methodology for measuring democratic legibility—for actually operationalizing what it would mean for citizens to be able to read their information environment—is, to my mind, one of the more interesting open problems in the field.
Media literacy asks, “Can you analyze this message well?” Democratic legibility asks, “Can you understand the system producing the conditions under which this message appears, circulates, and gains power?”
This newsletter is called The Frame. The name is intentional. In production, the frame is the boundary of what the camera shows—the choice, made by a human being and now often by artificial intelligence, about what gets in and what stays out. Every story is a frame. Every platform feed is a frame. Every editorial decision, every algorithmic ranking, every AI summary is a frame, whether or not anyone acknowledges it as such.
I want to write about what’s inside the frame and what’s been left out of it. About the research that’s changing how we understand information and democracy, and about the production realities that most researchers never see. About what artificial intelligence is doing to the journalism that Canadians and citizens of the world depend on, and what it means to teach the craft of media-making in a moment when the economics of that craft are collapsing. About democratic legibility—what it is, why it matters, and what it would take to actually improve it. I also want to write about how we produce media: techniques, tactics, tools, and tweaks. I’ll be drawing on my knowledge producing over two decades across different platforms, but also what current creators and I are also using. Because showing you how the media you consume gets made is key to helping to you better understand it. And, hopefully, change it for the better.
I’ll publish roughly every two weeks. The writing will be for anyone who takes media seriously, regardless of where they’re coming from—scholars who want their ideas tested against practice, journalists who want the research without the academic apparatus, students trying to figure out which questions are worth spending a career on, and producers who have always suspected that what they do for a living is connected to something larger than the broadcast.
The premise of The Frame is that those conversations belong together. That the people who have spent years deciding how, why, and what goes to air have something to say to the people who study what air time does to democracy. That the gap between making media and studying it is, itself, part of the problem.
I’ve been thinking about this for a long time. I’m glad to finally have somewhere to put it.
If any of this sounds worth reading—please subscribe, and tell someone who’d care.
If something I write is wrong, reply and tell me that, too. The best version of this work is the one that becomes a conversation.
—AJ