Voice cloning avatar technology is likely going to get much better and much more affordable very quickly. So, should we start using it for content creation? TELC thinks it’s a very interesting option to consider. Moreover, lecturers may want to use it too as they don't have to record anything themselves, which usually takes a lot of planning and valuable time. Having a mix of short-focused lecturer-voice-cloned avatar recordings and longer lecturer-‘real’ recordings (i.e. video conferencing and/or other lecturer recordings) might be a good mix to consider.  

Human-directed AI-enhanced content development could be the way to go, as AIs platforms such as ChatGPT have a lot to offer as a complementary source of learning content. In these sample outputs, I vetted and selected the content from GPT-4 on a range of topics; these videos could then be optionally used to support learners further. Additionally, platform APIs will help to speed up the video creation process.  

After listening to my AI voice a few times, I actually think it might be pretty good for non-native English-speaking students, as it is clearer and easier to understand. Obviously, the avatar could smile more, and I’m sure one day quite soon, it will be smiling more than me! 

Of course, closed captions can be generated too.