Beyond Text-Based AI : AI for Creative Application

Generative AI has changed how we create content, from text and code to images, videos, and music. While much of the conversation centers on text-based tools like ChatGPT, the scope of AI extends beyond writing. Non-text-based AI tools offer opportunities for creativity that can transform classroom experiences and teaching practices. For disciplines that don’t heavily focus on writing, these tools provide a reason to engage with AI technologies.

Non-Text-Based AI Tools can generate visual, audio, and multimedia content. These tools can be useful for creating instructional materials, enhancing classroom presentations, and sparking student creativity. Below are some prominent tools and their potential applications.

Image Generators

  • DALL-E: Now integrated into ChatGPT, this generator creates images from text prompts and descriptions.
  • Stable Diffusion: Customizable, open-source, professional tool. Ideal for those who want more control because it can run locally on personal computers.
  • Mage.space: Accessible and user-friendly diffusion model. Unlike Stable Diffusion, Mage is web-based making it more convenient without the technical set up.

Video Generators

  • Sora: This text to video generator is integrated into the paid subscription of ChatGPT.
  • Runway ML: AI assisted video generation and editing. Offers text-to-video, text-to-image, and image-to-video transformations.
  • Pictory.ai: Converts written scripts or slides into professional videos. Useful for flipped classrooms or repurposing existing content. Free trial available.

Voice and Sound Generators

  • NaturalReader: Generates realistic AI voices from text. Limited free voices many high-quality voices in paid version.
  • TTSMP3 – Free text to speech generator with multiple languages and MP3 downloads available. Character limit of 1000. 
  • Murf.AI: Offers realistic AI voiceovers and narrations in various languages. Perfect for faculty looking to add audio to their presentations or instructional videos. High quality, downloadable voices available for the paid version.
  • Soundraw: Enables users to generate royalty-free music. Students can explore music creation for multimedia projects or use it as background audio for podcasts and presentations.

A Closer Look: Prompt Variations Across Tools

To demonstrate how non-text-based AI tools differ in output, let’s use a single prompt across several platforms:

Prompt: “An underwater city with glowing coral reefs and futuristic architecture.”

Generated with DALL-E
Prompt: “An underwater city with glowing coral reefs and futuristic architecture.”
Generated with Runway ML
Prompt: “An underwater city with glowing coral reefs and futuristic architecture.”
Generated with Pictory AI
Prompt: “An underwater city with glowing coral reefs and futuristic architecture.”

Practical Tips for Using Non-Text-Based AI Tools

  1. Start Small: Experiment with one tool before introducing it into your teaching workflow. For instance, try DALL-E to create a single visual aid.
  2. Focus on Quality Over Quantity: Choose one or two tools that best suit your teaching style and needs.
  3. Engage Students: Teach students to use these tools to foster creativity and problem-solving skills.
  4. Stay Ethical: Be mindful of copyright and ethical considerations when using AI-generated content. Many tools offer clear licensing for their outputs.

Non-text-based AI tools open the door to possibilities for creative teaching and learning. Whether you’re designing visuals, producing videos, or creating audio, explore these platforms to help bring your ideas to life.