It’s even possible to imagine a pay-as-you-go model, in keeping with Willamon’s vision of a few hours released here and there throughout the year. In a variation on crowdsourcing, maybe Netflix could set subscription goals: Add so many new signups, and new episodes get produced. If something like that comes to pass, it would be seen as innovation. But it would also be returning to a business model that precedes television altogether. The serial was once a standby of movies, and of novels before that. After all, what are Dickens’ novels but collections of episodes that, when they first came out, were released serially?

The serial saw a seeming resurgence not that long ago with the popularity of Lost, a rare creative hit for broadcast TV. But picture Lost constrained only by the interest of viewers willing to pay more for more shows. Maybe viewers will resist changes to the formulas that have become so familiar that they’re no longer questioned. But the brilliant opportunity presented by television not limited by time or channel is that now technology has served up the chance to ask the question.

If Netflix or someone with similar resources is willing to take the risk, we can see if episodes or seasons are conventions that viewers still care about. And because we can all so easily vote with our dollars, the makers of TV will know quickly what’s working and what isn’t. In a way, platforms like Netflix make possible the same kind of iteration in TV that has driven computing tech forward so quickly in the internet age. If TV changes as much as the web has in the past decade, the stuff today’s iPad-addled kids watch by the time they’re adults might look nothing like what we call television at all.