People prefer to: Learn at their pace, whether for curiosity, hobby or work; Have learning resources available for quick access, e.g. on their devices; Be appreciated when they share their knowledge with others; Have a way to track what they learn, and refresh memory as needed; Have learning content conveyed to them in a non-judgmental, easy-going, friendly manner that respects that there can be different ways to do the same thing and people may choose their way after seeing possible choices.
Observation
Conclusion
People are more responsive to open-ended learning at their pace. And they like to pick out their preferred style of doing something, or of learning something, instead of someone deciding it for them. They do not want to be subconsciously comparing themselves with a human presenter, even if it is a video or still pictures of real humans. Why? Because if it is someone from their workplace, there is an automatic power dynamic between the person who knows and the person who's still learning!
Solution
Inspired by Wikihow, yet going further, the solution is: a hybrid of user-contributed wiki with visuals, content curation by select staff, and invitations to in-person/team learning sessions; great search engine and content ratings; relaxed and gently humorous in tone; tracks and rewards contributions; for wiki, replaces real humans in images with software-generated anonymized images; uses stills or GIFs of key how-to steps instead of longer videos; rewards completion of learning.
How would you stage or advertise your hack?
At various locations, a brainstorming wall with a large digital sign/screen of a barebones version of the solution, with marker pens and writing space around the digital sign for employees to make suggestions on what features etc they would want. Every night, I would add more elements into the initial barebones version, so that next day the employees see that "their suggestions" are getting incorporated and getting fleshed into the solution.
Reveal that incorporated suggestions as well as votes on prioritizing suggestions are earning coins/points... and thus start off employees with virtual coins/points that they can award for user contributions thereafter (and explain they shall earn coins/points for their own learning activities thereafter). One powerful aspect of this is that those who contribute content, i.e. "teach others," also end up widening their knowledge in the process, since the prospect of teaching automatically inspires one to learn in more depth than before.