I was curious about whether there was a difference in quality between content that I’d created through a process of scripting (physical writing), and doing the process in reverse, and using AI to transcribe a recording to generate a script.

Hypothesis:

Planning and scripting what I want to say, leads to a higher quality script than using audio transcription software to create a script from an “on the fly” recording.

Process:

I recorded a short video about something I have knowledge and experience about “what is learning design” using existing video software on my computer. I stripped out the audio file, and pulled the audio into an AI audio transcriber and summariser (parrot.ai). I used the free trial.

Was It Useful? What did I Learn?:

This was seriously impressive. I really enjoy writing, and find that it’s a useful process in terms of figuring out what I think, but it is very long, labour intensive, and ties me to my desk. What this opened up for me in terms of opportunities was doing my “thinking work” outside walking – where I think best, and then letting the ideas flow to camera, knowing that I could tear it up and start again. I think the expression of ideas is better, and more structured, than I would have got to via writing, certainly in a shorter amount of time.

The quality of transcription was good, I didn’t really need to correct any transcription errors, but when I read it back, there were certainly areas that I wanted to edit.

What I learnt was essentially that recording my thoughts to camera, at least for me, is a really efficient way of generating a script for something I know a lot about, once I have had some thinking and planning time. However, to get the best quality script, some editing for ideas and flow is needed. It also left me with the problem of how to take the transcript and create an engaging video or podcast.