Week 1 Discussion

👍 Q1. Which AI Assistant feature felt most intuitive or useful to you during the practice activities?

The most intuitive features are editing text and using the magic converter to change out blocks. The former is similar to other LLMs that I'm familiar with (Claude, ChatGPT, Copilot) while the latter is a straightforward way of instantaneously wrapping content in a new block format.

🚧 Q4. What challenges did you face when trying to get AI to produce the results you wanted?

The AI image generation feature was hit-and-miss for me. I think AI images have gotten very life-like, but the realistic one still have an "uncanny-valley" effect, while the more illustrative ones range in quality and often need significant prompt-tweaking.

As I typically need to follow brand standards, the illustrative ones, such as the vector-style, can diverge a bit too far from what is appropriate for our org. I find it easier to just use existing assets and tweak them in Illustrator. For realistic images, I have found it easier to just search on free picture sites or use the built-in media library. 

Personally, I haven't found a good use case for the image generator in my work at this point.