AI Images, Scenario Choices: ID Links 10/15/24

As I read online, I bookmark resources I find interesting and useful. I share these links about once a month here on my blog. This post includes links related to image generation and other AI resources, designing branching scenario choices, improving UX through font color, polite language research, disability images, and a document signing tool.

Image generation and AI resources

Generative AI Resources for All!

James Finder collected resources for learning about generative AI, including cheat sheets, newsletters, YouTube channels, and courses

StoryArtAI – AI Kid Stories Art Generator

StoryArtAI is an AI image generator designed specifically for creating illustrations for children’s books. This probably isn’t the most useful tool for workplace training or adult learners, but it might be useful for K-12 educators. The tool includes support for creating consistent characters and detailed prompting tips.

If you give FLUX1.1 a prompt like “IMG_1018.CR2” you get back images that are so very hard to tell they’re AI.

An odd tip from Twitter/X–use a file name with an extension for raw images as part of the prompt to generate realistic looking photos. I assume this is somehow prompting the model to look for those raw photos, but I wouldn’t have thought of trying this. I confirmed that prompting for an image file name (with .CR2 or other extensions) can generate some interesting images in Midjourney as well, although not always photorealistic images.

Free Image Vectorization

Convert PNG files to SVG. Useful for converting AI-generated PNG illustrations to SVGs for editing. Works best with flat illustrations without much detail.

Branching scenario choices

Game On: 6 tips for choice design in branching scenarios – Issuu

Miranda Verswijvelen’s article for Dirtyword magazine on choice design in branching scenarios. Lots of tips here based on designing game narratives and interactive fiction. I disagree with her point about not starting with the ideal path for writing (although she acknowledges that may make sense for beginners when you’re learning how to write scenarios). For game design, I think she’s right. For training design, there is typically an ideal path we want people to take. Good to read some thoughtful criticism though, and I love Miranda’s work (as you can probably tell by the number of quotes I highlighted to share below).

Excellent choice design will increase the engagement of your learners in the story, intrigue them about consequences and entice them to replay to check out other paths.

Choices in branching scenarios consist of three closely interconnected parts:

• Framing: the information the learner uses to make the choice
• Options: the possible choices
• Outcomes: what happens due to choosing one of the options.

Choices can offer diverse acceptable ways to achieve the same goal, giving learners opportunity to personalise the experience.

An extra path can also replace boring “try again” situations: the consequence shows your choice was not ideal, but you simply continue and get another chance further in the story to make a better informed, similar choice.

Clear and confined parameters help to make the choice feel integral to the context, while still leaving room for personal expression and emotional connection.

One of my favourite narrative designers, Jon Ingold from Inkle Studios, introduced the accept – reject – deflect model. For example, in a conversational choice where someone asks you a question, this could mean the following:

• Accept: continue the current conversation, e.g. simply answer the question
• Reject: react negatively or refuse to answer.
• Deflect: change the topic, e.g. ignore, bounce a question back or refocus attention

Taxonomy of Narrative Choices by Clara F-V

An interactive game in Twine showing some types of choices in interactive stories

Improving UX through font color

#BasicDesign — Never Use Pure Black in Typography

Explanation of why to avoid using pure black for text for online reading (which applies to websites and elearning). This article talks about how to pick shades of dark gray that are both readable and align with the color scheme.

White with code #FFFFFF has more striking brightness compared to pure black with code #000000. The brightness polarity requires the eye to work harder to read pure black text on a white background.

Font colors for ui/ux design instead of pure black #000000

Pure black text or backgrounds can cause eye strain and decrease readability. For more usable designs, use an alternative to #000000 such as these options.

Polite language research

The politeness effect: Pedagogical agents and learning outcomes

Using polite language in elearning improves learning outcomes.

The polite version yielded better learning outcomes, and the effect was amplified in learners who expressed a preference for indirect feedback, who had less computer experience, and who lacked engineering backgrounds. These results confirm the hypothesis that learners tend to respond to pedagogical agents as social actors, and suggest that research should focus less on the media in which agents are realized, and place more emphasis on the agent’s social intelligence.

Disability images

Disabled And Here: free stock images (with attribution)

Free stock images of people with disabilities, emphasizing people of color. While AI images are already usable for many purposes, I find that images of people with disabilities are still hard to generate. Images of actual disabled people are still useful. These are mostly photos, plus a few illustrations.

Document signing tool

BreezeDoc | Document signing is a breeze.

This is an alternative to DocuSign that looks like a good option for freelancers and small businesses. The free plan is limited to 2 documents per month, but the lifetime licenses are pretty cheap to increase that number.