So far TekLab has documented how teachers integrate generative AI to educate future journalists. The lecturer Jon Harman at the Volda University College (VUC) however, has started exploring the tools to teach students in other media fields like social media, video editing and animation. He is convinced that AI also contributes to the administration.
“This semester, I have been teaching storytelling with ChatGPT as a device. I told students to spend a day outside, observe the world and come back with prompts they can use to write a short story. Then they should use the chatbot to write a story with the same prompts and analyse the difference.”, Harman explains.
The prompt technique Harman used is called CLOSAT. This stands for
C: Description of CHARACTERS who could be used in a story
L: Interesting and visual LOCATION
O: Curious or evocative OBJECT
S: Loaded or Revealing SITUATION
A: Unusual or revealing ACT
T: THEME that intrigues, or you see in life
Harman pinpointed that the students came up with more innovative storytelling whilst ChatGPT presented a very well structure.
“This was the first experiment I did in class on how to use genAI in dramaturgy and storytelling and it shows that these tools can help media professionals in various ways. ChatGPT as a linguistic program tries to find a pattern and could help students to build a more profound structure", he concluded.
At the same time Harman challenged his students to use genAI tools to illustrate ideas.
“When doing a pitch for an idea, you cannot go out and film everything. You usually google images or mood board, but what if we use Dall-E or midjourney to mood things out?”, Harman asks.
The students were able to illustrate things that were in their head in an efficient way without consulting a designer. The previsualisation of things as an option was part of the bachelor level in media production at VUC.
Here are three pitches students created in collaboration with ChatGPT.
“I wanted to show students how we can use the tools in our work and stimulate them to think about their impact on their career in the future.”, Harman stated.
Therefore, he started discussions in an open plenum before and after the experiments. He did not yet implement it in the curricula but says that the playing and testing is the first step to integrate the education in genAI in a structured way with a red thread.
“I started thinking about the pedagogical aspect in June. The big trick is going through some processes bringing AI in classes. I think this prompt engineering in creative work is where I will focus some attention in the curriculum planning, I am a member of a couple of prompt engineering communities now seeking inspiration. ”, he says.
According to Harman there is a big problem with designing active learning. He criticises that many tasks and assessments are getting students to perform linguistically in a certain way and making it easy to cheat.
“I agree with Helen Beetham's perspective that our assessments should not just get students to regurgitate information recall or be performative, particularly in just a textual form - learning is deeper than this, as should be the evidencing of what has been learnt.”
Harman would like to see more debate on how AI may help teachers assess deep learning and knowledge with new tools.
“I come from a learning design methodology which often utilises an object orientated perspective on designing assessment and criterias. So I use prompts to assist formulating rubrics for assessment and phrasing of questions.“, Harman says.
Jon Philip Harman is a media producer and has been working in TV production and online media in the UK.
He played around with AI in the media space such as video tools, Dall-E and chatGPT and used midjourney already successfully for a pitch in animation. By doing that he skipped months of waiting for an illustrator to be done with the prework.
He also conducted gave students insight into the creative industries in an self produced podcast which he uses as a teaching resource.
“I therefore did an episode with Chat GPT utilising an AI generated voice and picture. Incidentally, I use AI tool Descript to edit the podcast which speeds up my workflow from standard editing workflow.”, Harman says.
What he wishes from AI in the future is to be integrated in post-production tools, video cameras and so on.
“I would like it to reduce a workflow pipeline, but not to be an automated process like on Canva. It could simplify the production and makes the life of a professional easier in the way that I don’t have to go through three layers of profession.”
What he means is that media professionals often are dependent on how fast other professionals can do their part before they can continue their own work in the production. This evolution however will eradicate many jobs, particularly in motion design. Harman understand the danger of the AI evolution but also sees a necessity to improve administrative tasks.
“There is a lot of wasted bureaucracy in the way contracting and payments are done. Also, the way we do analytics, or the workflow of bureaucracy could be improved with AI.” Harman states and continues:
“All narratives in higher education are about cheating, but no one is talking about radically transforming studying administration and the behind the scenes. AI could simplify their needs to do research, but also the industry must think about it.”
Harman challenges them to find and use the tools that are already given. Outlook for example has a bunch of AI tools build in, but according to Harman people are struggling in using them for such purposes.
