Executive summary
Universities do not have a video problem. They have a production problem. Most institutions can already record every lecture they need. Very few can turn those recordings into finished, consistent, accessible courseware at the scale of a programme — and the gap is widening as regulators, students and competitors raise the bar.
This paper looks at how lecture video is produced today, what it really costs, why adding AI tools has not solved the problem, and what a different approach — Academic Production Intelligence — looks like in practice.
1. The bottleneck is people and handoffs, not cameras. A single finished lecture typically passes
through eight to twelve roles and four to six vendors or tools. Each handoff adds delay, cost and a chance for
inconsistency.
2. The traditional cost structure does not scale. At common editing ratios of 6 to 10 editor-hours per
finished hour, a 400-hour programme needs 2,400 to 4,000 editing hours — more than one full-time editor can
deliver in a year — before captions, design or review are counted.
3. AI tools have added vendors, not removed work. Transcription, captioning, clean-up, image
generation, video generation, stock libraries and editing each arrive as a separate product, contract, data
review and file format. Someone still has to assemble them, and nobody owns the result.
4. Generic AI does not understand teaching. Video tools know about cuts and motion. They do not know
what is being taught, how a discipline explains itself, or when a visual would interrupt a derivation on the
board.
5. A single academic production pipeline changes the economics. Producing 400 finished hours on
Lecture Studio costs about $7,200–24,000 depending on tier, with one vendor, one price per
lecture, one approval step and no new hardware or re-recording.
The position in one line. Lecture capture records the teaching. The LMS delivers it. Lecture Studio
produces it — turning faculty recordings into finished digital courseware, with faculty as the final word on
everything students see.
What is inside
15 pages, 7 exhibits, an About Lecture Studio page, sources and assumptions listed.
- The problem: recordings are not courseware
- How lecture video is produced today
- Too many stakeholders, no single owner
- What it really costs
- Why AI tools have not fixed it
- What Academic Production Intelligence means
- How Lecture Studio works
- Clarity and ease of deployment
- What is built today, and what is next
- A checklist for evaluating any production approach