Lucas, a blind individual, demonstrates JAWS Commands using AI Descriptions and how to effectively use ChatGPT with DALL-E to generate and describe images. Starting with ChatGPT, Lucas opens the platform, which places him directly in the edit box. To navigate out, he presses the “up arrow” key. He then maximizes the screen with “Windows + up arrow,” enhancing readability for JAWS.
Next, Lucas demonstrates how to locate the DALL-E AI Description image generator by accessing the links list using Jaws Commands with “Caps Lock + F7.” He navigates to the search box using the “E” key and types “DALL-E” in all capitals, with the period included. After pressing “Enter,” he tabs through the search results until he finds the relevant link, then presses “Enter” again to open it.
Once inside the image generator, Lucas describes the process of inputting a prompt. He emphasizes the importance of being specific, including details about the setting, characters, and art style. For instance, he generates an image of a wizard casting a spell on a cliff-side in a realistic art style. By pressing “Enter,” the image generation process begins, typically producing two images by default.
To download an image, Lucas uses “Shift + B” to find the “Show Image” button, followed by pressing “Enter” and then down arrowing to the “Download Image” button. He encounters a warning about the dialog not being focused, but the download proceeds.
For image descriptions, Lucas follows a similar process to find the “Image Description Assistant” link. After inputting his request and attaching the image file, he waits for the description to generate. To review the description, Lucas uses “Shift + G” to jump to the top of the graphic, then down arrows to read it. This methodical approach allows Lucas to create detailed images and receive comprehensive descriptions efficiently.