Pan and Expand: Midjourney's New Visual Tool
Pan and Expand: Midjourney's New Visual Tool
Go beyond the boundaries of possibility
Total reading time is about 3 minutes.
This edition is sponsored by Salad
The lowest GPU prices for inference at scale on Salad
Tired of GPU shortages, expensive cloud bills and an over-reliance on A100s/H100s?
Access 10k+ consumer-grade GPUs starting at $0.10/hour and save up to 90% of your cloud cost on SaladCloud. Our recent stable diffusion benchmark generated 9M+ images in 24 hours for just $1872 - the lowest price anywhere in the market.
Get a demo to see how you can get more inferences per dollar for your use case.
Midjourney Pans Out
Midjourney rolled out the “Pan” feature this week to add more options for users to ‘fill in’ the areas on either side of an image or up and down.
You can Pan to the right or left or up and down. Each 'Pan’ results in 4 images with slightly different panned areas. You can see the effect in the example below.
The original image is in the center, inside the white rectangle.
I added color effects to the panned areas to help show the seamless expansion of the image using this feature:
Here are a few more ‘Pan’ examples below the original image:
Now, you can combine the ‘Zoom’ feature with ‘Pan’ to broaden your storytelling:
Why is this a big deal?
Generative AI technology is improving every week, and it feels like we’re light-years away from the results one year ago. Our creative options are continually growing with a single mouse click!
Look at the difference between the original image and the last one - which one tells a more descriptive story?
🔮AI News
This week was slower than usual for generative AI news, but there were a few major stories:
OpenAI Announcements
Code Interpreter is available to all ChatGPT Plus users. Run code, edit files, analyze data, create charts, and more.
You’ll need to turn it on in the settings, as shown in the image below:
OpenAI also announced the general availability of GPT-4, its latest text-generating model, through its API.
Access to the GPT-4 API will be limited to developers with a history of successful payments, but the company plans to open up access to new developers in the future.
Additionally, older models in the Completions API will be deprecated by January of next year.
Advertise your AI job, event, or tool with Visually AI’s community of over 5K subscribers:
🚀 This Week’s AI Tools
Photoleap by Lightricks: AI-powered mobile photo editing app with text-to-image, sketch-to-image, and animation, including a fun Twitter bot that generates images if you comment @photoleap_ai. (link)
Leonardo: Full-service AI image generation platform - introduced an updated version of their Canvas editor, with inpainting and generative fill, like Photoshop. (link)
Sivi: AI-powered graphic design generator creates customized marketing visuals in seconds. (link)
Zealous: AI quickly turns recordings into social media & marketing content. (link)
Skybox AI: Blockade Labs announced the launch of Sykybox AI Model V2, with improvements in clarity, details, and accuracy in text-to-360° virtual worlds. (link)
🎁 Get it free: The AI Visual Creator’s Toolkit
Boost your content with my all-in-one, free visual AI toolkit!
Access AI-powered tools for AI-generated images, image editing, and more:
Thanks for reading and have a creative week!
Heather, this nails it:
"it feels like we’re light-years away from the results one year ago. Our creative options are continually growing with a single mouse click!"
Indeed. One year from now is very likely to have a great deal more change than the last year, too, due to accelerating returns from existing technology. This iteration will, in turn, make the next one better, and so on. We're at some kind of liftoff for visual arts, and it is amazing!