Why a person still ends up in the booth
The projecting part of projection is already automated. Digital cinema servers hold the features, a playlist cues trailers, lights and masking, and the show starts on a timer. That change happened with digital projection, not with AI. What keeps people in the job is everything the playlist cannot touch: inspecting and splicing prints where 35mm or 70mm still runs, swapping lamps, cleaning optics, and getting a stalled show back on screen while a full house waits in the dark.
The scale of the occupation reflects that older shift. BLS counts about 1,480 projectionist jobs in the United States, with median pay of $38,270, and projects a 6.2% decline between 2025 and 2035 (BLS, 2025). So this is a small, slowly shrinking role. The shrinking came from automated playback and multiplex staffing, which folded booth duties into manager, usher and technician jobs. Fewer entry-level booth hires is the honest story here, not software taking over a trade.
Physical work is the other reason. The robotics panel on this page puts most of the duties in the hands-on column and places the job in the mobile robots tier. A model can read a server error log. It cannot climb a ladder in a hot booth, seat a xenon lamp, re-tension a platter or re-rack a print that jumped a sprocket. That gap is hardware, not intelligence.
What software runs, what it assists, and what stays manual
Automated cinema systems handle the scheduled side of the work on their own. Building and triggering a playlist, and matching sound, lighting and curtain cues to the start of a feature, are jobs a server does without supervision once they are set up. Our share of task time in that group: 0%. You can read how we measure that share on the coverage method page.
A second slice is assisted rather than automated. Watching alarms for a dropped signal or a failed ingest, and keeping records of screenings, lamp hours and projector service, are faster with monitoring software than with a clipboard, but someone still decides what the alert means. The assisted share reads 5%.
The rest sits with people: 95% of task time. That is the manual trade. Inspecting a print for damage and splicing it, replacing lamps and cleaning lenses and ports, setting focus and framing for a specific screen, and troubleshooting a failure during a paid show with an audience already seated. The cost panel above is also part of the picture: the software side is cheap, but it only covers the scheduled tasks, so the staffing cost does not disappear with it.
Good to know: automation in cinema projection removed booth hours before modern AI arrived, which is why the task split looks the way it does.
What has actually been tested
Not much, and that matters. Our evidence grade for this occupation is D. At that grade there is no direct, published test of an AI system against a working projectionist on this job’s real tasks, so we publish no quality parity number at all. Guessing one would be worse than leaving it blank.
What would settle it is specific: a measured trial of automated fault detection and recovery across a run of live shows, compared against a trained operator on the same equipment; or a study of print handling and splice quality by machine versus hand. Broad studies of generative AI in film and television production look at other roles, such as editing, visual effects and motion graphics, and do not test booth work. Until something closer exists, the honest read is an untested occupation with a small automated core. How grades are assigned is set out on the quality parity method page, and the full scoring approach is on the methodology page.
When the picture could change
Most likely after 2043 (8 in 10 of our scenarios). What that window measures is explained on the replacement year method page.
Two things could pull it earlier. First, further centralization: one technician covering several sites remotely, with automated monitoring handling the routine faults, cuts the hours needed per screen. Second, cheaper service robotics. If mobile machines can reach, lift and swap parts reliably in cramped booths, the physical blocker weakens.
Two things hold it back. Projection rooms are unstandardized, with mixed generations of equipment, different screens and formats, and failures that happen once a year and have to be fixed in minutes. And the downside of a mistake is immediate and public: a dark auditorium, refunds, and a manager with no backup plan. Venues pay for a person because a person can improvise. If you want to see how that compares with occupations where the automated share is much larger, the most exposed jobs list is the place to look.
How to stay needed in projection
Lean into the work that stays manual. Fault recovery under time pressure is the single most valuable thing you do, so build a record of it. Film handling, including inspection, splicing and platter or reel-to-reel work, is scarce and is what repertory houses, festivals and archives pay for. Picture and sound setup by eye and ear, including focus, framing, masking and room calibration, also stays with people.
Two skills extend that. Learn the digital side properly: servers, key delivery messages, ingest, network playback and the diagnostics that come with them. Then learn the wider room, which means audio systems, rigging, lamps and basic electrical safety. That is the path from booth to technician.
Nearby jobs worth comparing: ushers, lobby attendants and ticket takers, amusement and recreation attendants, and audiovisual equipment installers and repairers. You can put any two of them side by side with the compare tool, read the rest of the entertainment attendants family, or see how the wider arts and entertainment sector is scored.