AI assisted content creation is an exciting new approach driven by algorithms and models like GPT-3, DALL-E, Stable Diffusion and more. At CodeArt we've been experimenting how to use these technologies to enhance the editorial process. In this second part of the blog series on the topic I'll share a video of a prototype integration to Optimizely (Episerver) CMS of content creation using AI for both images and text.
This is part 2 in a blog series about AI assisted content creation. You can read part 1 here.
Over the last 4-5 months, in a collaboration with the great Grzegorz Wiecheć, I've been working on a prototype integration of both GPT-3 (AI assisted text creation) and DALL-E / StableDiffusion (AI assisted image creation) for Optimizely CMS.
At Opticon in San Diego in October 2022, I was invited to give a showcase demo which received a lot of interest.
To make it available to many more, I made a quick youtube recording of the demo:
Essentially, there are two key parts to the integration:
1) An attribute you can attach to any text property (XHTML or string) to enable a small button next to the field, that will send it to GPT-3 and return the output.
This could potentially also be used to provide other AI functionalities to that property - like auto-generating meta-data for an image (using cognitive services) or auto-suggesting a taxonomy-category, or perhaps extracting a list of Named Entities in the text.
The other is something as simple as a "Generate" button on the media folder, allowing you to easily get prompt and image picker that can work with image generation services like DALL-E or Stable Diffusion through their API.
This is currently just a basic prototype, but we are exploring different ways to make this available so more people can play around with it - and start enhancing their content today.
Feel free to drop a comment below - or reach out to me directly if you have a use for this and would like to learn more.