Using AI has become a daily practice for many learning design professionals. I sometimes weekly review some of the AI tools and browser integrations that look promising for learning designers to perhaps try out. In today’s post, I’ll again share the Descript AI tool below. I did mention this tool in a previous post and since that time it has made major updates.
In January 2023, I tried out the AI Descript tool for multimedia creation using different generated AI images and AI voices. The intention continues to be an exploration and investigation of how AI is continuing to evolve in the learning design space using multimedia for audio and image creation.
Hi @WJRyan ,
Thank you for asking about the process and prompt. The Descript AI works similar to PowerPoint but has AI built in features like text enhancement and AI image search, and many features for accessibility like CC, transcripts, speed adjustments, and includes an audio only designation and allows for sharing, commenting plus more. It is a download application and publishes to their website upon request. The design production layout was intuitive and the editing timeline was functional with basic multi-media editing features.
I noticed previously DALLE 2 was the AI imager generator, no longer available. In it’s place is another type of AI image search which I found frustrating to use and seemed to have limited results and had a pay for play option which didn’t appear to be enabled in January unless I missed it.
My process wasn’t from design template but rather I needed to quickly generate a presentation so I choose to create the project from the slide deck for a presentation itself.
Prompt Generation:
From a short, detailed prompt Chat GPT generated a response that I chose to work with. Prompt: Create three wellbeing recommendations using AI for in the metaverse to help with employee stress and anxiety .
From that response, I ended up with 11 slides. I entered most of the text from ChatGPT and then added AI images. After I completing that, I began a quick review and made very few contextual edits.
One of primary edits was to choose an AI voice to read the slides. I choose a few different voices to add interest.
In addition, to keep interest, I choose to upload an audio AI file to provide some background ambient sound. The presentation was close to exactly 2 minutes in length. The creation of the entire project was estimated at 2 hours.
As I became more familiar with the features of the AI, I believe if I used it more frequently the EST for a deliverable would significantly decrease of course depending upon the length of the slides and if uploaded media were to be added this would be additional time to be factored in.
