by Sean Sweeney October 30, 2023 2 min read 1 Comment
By Sean Sweeney, MS M.Ed., CCC-SLP
You have probably heard chatter about the potential and possible pitfalls of AI in educational contexts. This month: a simple way to harness the possibilities of AI image generation.
Generative AI has been around for many years but has only been recently been made widely available to the public via a variety of webtools. This kind of artificial intelligence is called “generative” for its ability to use growing neural networks to create content, such as novel text, imagery or even music and video. ChatGPT is a recent tool which responds to prompts to create text, and I recently blogged about its usefulness in generating problem solving scenarios–also useful for targeting story elements.
Along with ChatGPT, Microsoft’s OpenAI offers DALL-E, an image generator which is able to create pictures from “natural language,” as opposed to code. This is most easily accessed through the Bing Image Creator, a simple tool that develops images from scratch (not search)!
Check out this image below, generated with the prompt “A farm with a cow and pigs. The cow is refusing to be milked by the farmer:”
Image Created in Bing Image Creator
Note the lesson possibilities here, as what you have is perfect alignment with Story Grammar Marker®: Characters, Setting, and Kick-Off, with inherent Feelings and the opportunity to discuss Mental States and Actions or other story elements.
A prompt to Bing Image Creator generally gives you 4 results, so you (or your students) can decide which one best depicts your story. You can also cater your image to a certain style of art, such as anime, clip art, or line drawings.
Bing Image Creator Offers You Up to Four Images
This style could then become a great coloring, as well as a story analysis activity! As you can see, the story options are virtually limitless with this kind of tool. Be aware that the Image Creator does less well with proper nouns, such as specific celebrity or product/business names, so keeping it generic is a good idea. Historical figures are a possibility for aligning with curriculum topics, however!
For more on AI tools useful in education, check out Ditch that Textbook’s terrific overview.
Happy creating, and Happy Halloween!
Sean Sweeney, MS, MEd, CCC-SLP, is a speech-language pathologist and technology specialist working in private practice at the Ely Center in Needham, MA, and as a clinical supervisor at Boston University. He consults with local and national organizations on technology integration in speech and language interventions. His blog, SpeechTechie (www.speechtechie.com), looks at technology “through a language lens.” Contact him at sean@speechtechie.com.
Comments will be approved before showing up.
Tania Fajardo, M.A., CCC-SLP
December 23, 2023
Hi! Yes! I recently discovered Adobe Firefly (and though my daughters said I was way behind finding out about AI) I saw a great new activity to try with articulation groups. Once they are at the phrase or sentence level, I have them generate their targets and type as they speak. If they have made errors, we may get outrageous visuals! Another variation is working on language concepts. I see lots of possibilities, but you are right about the storytelling potential here. It goes hand in hand with the Six Second Stories methodology, too. Good post! Thanks!