Week 2 - Group 2 Discussion

I think the part that took the longest for me is overcoming my own perfectionism. I knew that this was relatively low-stakes and yet I probably spent way too much time tweaking and adjusting the course.

The advantage to being diligent was that it allowed me to really dig into Rise 360, something I needed a crash course in anyway, so this was very much two goals achieved at once. I really dug into the AI features because our team just upgraded at the start of the cohort, so I wanted to understand all we could now do.

With that, here is my course.

Major Changes Made

1. AI Image Generation. I went through and changed all the images the AI Assistant came up with and used my customized AI generated image prompt for the whole course (with the exception of the lesson headers). I wanted to have cohesion throughout the course as much as possible by using the same style.
2. Updated Tone. I kept wavering on the tone I wanted to strike throughout the course and I decided to keep it "straightforward" as much as possible. So I updated blocks of text wherever I found it got too wordy to make it shorter and then the tone straightforward. 
3. Instant Convert Blocks. There were several blocks the AI Assistant came up with that I found didn't fit the pedagogical narrative of the lesson, so I converted them to a block that fit better. I understood why the AI selected them, but I opted to go with something I felt more fitting.

Most Valuable Feature/Capability

While I spent a bulk of my time converting text tone/simplifying, so it would be easy to say that feature would be the most valuable, I would actually say the initial set up process was the most valuable capability.

Converting text is great, but for the purposes of this exercise, I wasn't paying as much attention to the quality of the output. I would check it and it was fairly accurate to what I felt would be appropriate to the lesson, but in real life, I would be writing everything myself.

The set up process, however, that tends to be the biggest mental load in the whole course creation process. Being able to put in all the source materials, create guidelines, guardrails, and edit the output so that the AI Assistant does the bulk of the heavy lifting? That is the real MVP of the whole process here.

We have a lot of courses where material was copy/pasted from an internal wiki into Rise 360 for the sake of creating training material quickly. But with the AI Assistant, we can take that source material, plug it in, and actually create a meaningful course that is geared towards active learning rather than wiki regurgitation. I see us doing a massive "retooling" of our course library in the near future.

Most Challenging Moments

Beyond my own perfectionism getting in the way, it's the usually issue of trying to get the image generation just right. Generation has come a long way from the early days and hallucinating has gone way down, but there were a few errant hands here and there that I needed to fix before finalizing the course. Admittedly, I may have missed a few.

The other issue with image generation was still having issues with consistency in styles. I wanted a specific look, but wouldn't always get it. I know that was my problem with not being more specific and fiddling with my terms/guardrails a bit more.

Final Thoughts

Overall, I found this cohort extremely useful for my daily work and see it helping other members of my team that work in Rise. The AI Assistant feature will cut down a lot of moments where work can get blocked/slowed down and it will speed up the course creation process.

The only critique is the usual one is getting too reliant on the AI because I can easily see that happening and I don't want to delegate too much of the work out to the tool. I still want to provide enough of my own oversight.