For most of its history, screen work in Canada has been a service industry. Vancouver, Toronto and Montreal have spent decades standing in for New York, Boston, Chicago and Seattle, and that arrangement trained a generation of performers to think of the audition as a private, carefully lit transaction: a reader, a camera, a coverage sheet, and a link sent to a casting director who may never be in the same room.
The feed has rewritten the shape of that transaction. Casting notices increasingly ask for a short vertical clip — a dance, a look, a fifteen-second reaction — shot on a phone, in a bedroom, in one continuous take. For performers who were never asked to hold a frame alone, that is a new craft requirement. For the people making the work, it is a new kind of production line.
The Grammar of Twelve Seconds
A vertical short is not a shortened film; it is a different grammar. The face has to read at thumbnail size. The first second has to do the work of a title sequence, because a thumb decides faster than an eye. One cut is enough and often one cut too many. Blocking that looks ordinary on a monitor can read as fidgeting in 9:16, and a performance that feels too large in a room can be exactly right on a phone.
That is why formats people already share are so useful as a starting point. A point-and-swap bedroom dance, a deadpan white-suit dance in an empty room, a kitchen argument, a dog crew on a black stage: each is a small stage with its own rules, and each teaches something about timing that a general audition never does.
What a Short Actually Costs
Short-form production is not cheap. Even a modest concept needs a location, light, a person to hold the camera, someone to cut it, and a soundtrack that has been cleared. For an independent producer in Halifax or Hamilton, that is a day of favours and a week of scheduling.
This is where the tooling has moved fastest, and where the vocabulary needs care. The interesting question is not whether AI can generate a video — it clearly can — but what a tool is willing to refuse. The performers and crews most exposed to this technology are not worried about the output looking unreal. They are worried about being put into footage they never agreed to appear in.
Where the Line Gets Drawn
One example of that boundary-making is Rapsona, a photo-to-video tool built on a fixed set of templates rather than open-ended prompting. You upload one portrait, or two for a duo, choose a format, and receive a twelve-second vertical clip with an original beat and English rap vocals generated alongside the picture. The style library is fixed — a handful of templates tied to dance trends currently circulating — and the setting, wardrobe, camera move, tempo and lyrics are set by the template, while the face match, mouth timing, voice, hand movement and background detail vary from take to take.
At the time of writing, a single video costs $5.99, a failed generation returns the credit, results typically arrive within half an hour, and finished files are deleted after seven days unless you download them. Photos and videos are not used to train AI models. What it will not do is clone a voice, swap a face into existing footage, or lip-sync a song you upload — which is precisely the line most working performers want drawn.
Consent Is the Story
The ethical questions are not decorative. Every template requires an adult who has agreed to appear; public figures and minors are refused outright, and the tool asks users to label the output as AI-generated when they share it. Personal and lawful commercial use is allowed with the agreement of the person pictured. Whether or not these rules hold up in every case, the direction is right: the consent of the performer, the disclosing of the synthetic frame, and the refusal of the two things that would make the format predatory.
For Performers and Crews
None of this replaces the work. A twelve-second template is a distribution channel, not a screen performance, and the actors worth casting are still the ones who can hold a scene for four minutes. But the economics have shifted in useful ways. Test footage, pitch visuals and mood boards that once required a favour from a friend with a camera can now be produced for the price of lunch — and the firms entering the vertical space also tend to ship the unglamorous pieces, such as a free subtitle tool that exports SRT and VTT files, which matters more in an industry where most video is watched on mute than anyone admits.
There is a harder point underneath. Canadian crews have spent decades proving they can make anything look like somewhere else. The short-form feed asks something less forgiving: it asks whether the person on screen can carry a frame alone and whether the production behind them is honest about how it was made. That is a craft question, and it will not be automated away.
The Feed Is the New Screen Test
Canada will keep standing in for other cities, and the soundstage will keep its place. But the first frame of a career is now increasingly shot vertically, in a bedroom, on a phone, and cut in one take. The industry’s job is to make that frame honest: consented, labelled, and a little better than it needs to be.