This is one of the new AI Design Assistant tools in Blackboard that have been made available to staff from January 2024. Their use was recommended by the Educational Development Committee in November and approved by Education Committee in December 2023.
Purpose
Use of this tool is entirely optional. Anthology have recently expanded the ways that you can add images into Blackboard courses. As well as the traditional route of uploading an image from your computer, in October 2023 they added the ability to search for a royalty-free image on the Unsplash site. This AI tool adds a third option – to generate a new image from a description you provide, using the DALL-E (updated to Version 3) tool (via Azure OpenAI). You don’t need to create an account to use this and it does not cost you anything. Once inserted into the course it behaves like any other image.
It is only visible to staff with any one of the following roles in the course:
- Instructor
- Module Administrator
- Course Builder
- Teaching Assistant
Video demonstration
Step-by-Step guide
Finding the tool
This tool appears when you click the Insert Image button in the standard Blackboard Text Editor:
This opens the Insert Image dialogue. By default it offers you the option to Upload from Device, but you can highlight the Image Source list to select an alternative method:
You should select the option Generate images to use the new AI Design Assistant.
Getting good results
The tool will try to generate a potential image as soon as it is launched (based on the Course Title), but for best results you will want to Describe the image and ask it to generate again. Previously, this was 4 images, however to streamline and speed up the process only one image is now created at a time.
Description
Use the description to describe the style and content of the image you want – e.g.
- a photorealistic image of a small old man holding a large balloon
- a cartoon showing a blue whale floating in the sky
- a dusty laboratory desk, covered in flasks
Generate
Use the Generate button to apply your changes. As before the tool generates potential matching images for you to choose from.
If you can’t find a suitable image, press Cancel to close the dialogue.
Note that many of the images generated may reflect biases in the algorithms and/or training data (e.g if you ask for an image of a nurse it is very likely that the person generated will be female), so always check them carefully before use.
Selecting the image to use
Each time you click the Generate button the images generated are added to the image bar. When more than 4 images have been created you will see Previous and Next arrows appear at either side of the images and coloured dots beneath the images to indicate which page you are on (left image). If you like one of the generated images, highlight it (right image) and the press Next.
Edit the image
As with images you upload or select from Unsplash, you can then use the editing tools in Blackboard to resize or zoom in on a particular region of the image
Aspect Ratio
By default the image will appear as a square. If you want to change this, you can use alter the Aspect Ratio to crop the picture to create a portrait or landscape-shape. The select offers some standard width to height ratios.
You can move the selection up or down or left to right, to centre it on the area you are most interested in.
Zoom
You can zoom in or out of the image using the slider tool below it. This can be helpful to ensure the focus is on the correct part of the image. Note however that images created in Blackboard via generative AI are typically no larger than 1000px square, so will rapidly pixellate if you zoom in too far!
Edit the File Options
When you are happy with the appearance, select the Next button to switch to the Edit File Options dialogue. This is the last step and has three sections:
Display Name
This is the name that the file will be given in the course Content System. It is not normally displayed in the interface, but if you give it a sensible one (in our example I have used dusty_laboratory_image.jpg) this may help you find it again. If you allow students to download it (see below), they’ll also thank you for that.
Accessibility
You have two accessibility options. You can either provide Alternative text for the image (which is usually the best thing to do) or if the image really doesn’t add anything to the materials or convey any key information, then you could Mark the image as decorative (which means screen-readers will skip it).
When writing alternative text, you don’t need to use the word image or picture, and often the text you entered into the image prompt earlier is a good candidate – e.g. a dusty laboratory desk, covered in flasks.
File Options
This setting determines what students on your course can do with the image. If it is just there for decoration, you might want to limit it to View only. If it is important that they can make use of it, then use an option that permits them to Download it.
Save
Selecting Save inserts your image onto the page.
Remember to save the text using the Save button below the content editor too!
Acknowledging AI use
When the tools were approved at Education Committee a student rep suggested that staff should acknowledge their use in the course, just as we ask students to acknowledge the use of generative AI in their assignments. This was seen as good practice and the recommendation approved.
Our AI Design Assistant: Acknowledging Use page provides some suggestions.