Creativity AI #29: Weekly update for AI Art
News about AI art, Updates from Geeky Curiosity, Prompt Play, Inspiration
Creativity AI is a free weekly publication that brings together interesting articles and recent developments in AI art and creativity.
You can also read this newsletter on the website.
News about AI art
Midjourney’s Relaxed mode situation is becoming increasingly complicated, as the top 1% of users are heavily utilizing it, resulting in resource and financial strain to Midjourney.
Originally designed for excess server capacity, Relaxed mode may now experience increased waiting times, more variable delays, or adjusted concurrent job limits as the excess capacity is no longer available.
Midjourney's video model currently trains with 5-10 times less computing resources than tech giants like Google. Still, the developers are also actively exploring partnerships with other AI labs to boost their resources.
Midjourney is optimizing its servers specifically for faster video generation and reduced costs, with exciting new video features on the way, including a video Upscaler and improved artifact reduction.
Future video updates include start-to-end frame generation, enabling smooth animation between images and looping videos, as well as "Turbo mode" for faster video generation.
Midjourney V 7.1 is currently in development, though the team faces some challenges in combining video and image models into a single release.
Version 8 has already entered the early training stage.
Two secret projects are nearing the announcement stage, although details remain under wraps for now.
With their 60-person team, Midjourney expects to roll out more frequent releases in the future.
This summer, KlingAI released their "SurfSurf" effect, which allows any character to surf on the wave. It's quite fun to play with. The surfing animations also have background music. Check out what I have created.
Midjourney images:
KlingAI SurfSurf Effects:
Ideogram 3.0 joins Adobe Firefly's AI ecosystem alongside OpenAI, Google, and Black Forest Labs models. This integration brings Ideogram's text-to-image capabilities directly into Adobe's creative suite and mobile apps, giving the users access to a variety of AI generation tools within their familiar workflow.
Napkin AI just rolled out something that's quite practical and useful. This free infographic tool that I've been using since its early days now has a "Custom Generation" feature.
You can specify your visual type, pick orientation, or keep things closer to your original text when creating graphs and charts. Perfect timing for my upcoming presentation deck. Watch their intro video on YouTube.
Runway Gen-4 has updated its image references, and the object consistency is now quite powerful, although the reference text may not always be rendered correctly. For character consistency aspect, they're beating Midjourney right now.
Midjourney's Omni-Reference is stuck in development with just one image support, while Runway supports up to three reference images. Big difference for consistent video generation.
Update from Geeky Curiosity
Your curious mind wanted. Help shape Geeky Curiosity.
I'd love to hear what's sparking your curiosity these days and how I can make this newsletter even better for you.
Your thoughts help shape every article I write, so please take a moment to share what's on your mind.
Click here to share your thoughts →
I’ve updated the Midjourney Style Resources #8 to V7 with some stunning golden visuals.
Prompt Play
Ever wondered if AI can actually make you hungry? I decided to put three popular image generators through their paces with some mouth-watering food prompts.
Today's Prompt Play is simple. I fed the same two prompts to ChatGPT -4o, Midjourney v7, and Ideogram v3 to create an Italian pasta dish with fresh ingredients, and whip up a tropical fruit salad with various fruits.
ChatGPT followed the prompts surprisingly well. The pasta looked decent, and the ingredients were clearly visible. But when it came to the fruit salad, those pineapple chunks looked like mangoes. But ChatGPT shows you a cut of pineapple with its skin outside the bowl.
Generate an image of a classic Italian pasta dish with fresh ingredients. Show the finished pasta on a rustic wooden table, surrounded by tomatoes, basil, garlic, and Parmesan cheese. Use natural lighting and a warm, inviting color palette.
Create a vibrant image of a tropical fruit salad served in a modern kitchen. Include colorful fruits like mango, pineapple, kiwi, and strawberries in a glass bowl, with some fruit pieces and mint leaves scattered around. The background should be bright and cheerful.
Midjourney v7 images looked so photorealistic I wanted to reach through my screen and grab a fork. The pasta had that perfect al dente shine, complete with visible cheese and herbs. Even better, the fruit salad actually looked like something you'd find at a high-end resort buffet. But again, it is harder to differentiate between pineapple and mango inside the bowl.
classic Italian pasta dish, fresh ingredients, finished pasta served on a rustic wooden table, surrounded by ripe tomatoes, fresh basil, garlic cloves, and Parmesan cheese, natural lighting, warm and inviting color palette, high detail, food photography style --ar 16:9 --profile i6fyyx5 --v 7
vibrant tropical fruit salad in a glass bowl, modern kitchen setting, colorful fruits including mango, pineapple, kiwi, and strawberries, some fruit pieces and fresh mint leaves scattered around, bright and cheerful background, vivid colors, natural lighting, high detail, food photography style --ar 16:9 --profile i6fyyx5 --v 7
Ideogram V3.0 landed somewhere in the middle. The realism factor was definitely much better than ChatGPT's. But it shared the same pineapple confusion issue. If I showed you those yellow chunks without context, you'd probably guess mango, too.
classic Italian pasta dish, fresh ingredients, finished pasta served on a rustic wooden table, surrounded by ripe tomatoes, fresh basil, garlic cloves, and Parmesan cheese, natural lighting, warm and inviting color palette, high detail, food photography style (Magic prompt on)
vibrant tropical fruit salad in a glass bowl, modern kitchen setting, colorful fruits including mango, pineapple, kiwi, and strawberries, some fruit pieces and fresh mint leaves scattered around, bright and cheerful background, vivid colors, natural lighting, high detail, food photography style
Conclusion: For food bloggers or recipe creators, Midjourney v7 is clearly the winner here. But the other two AI tools are great.
Now I'm curious about what happens when we test them on more complex dishes!
Inspiration
Sometimes, the best art comes from complete accidents. I experimented with chaotic Midjourney prompts by cramming visual keywords together without structure (aka "keyword salad")
The results? Gorgeous abstract collages. The bot works hard to keep things pretty despite the state of chaos.
Embrace the chaos.
A colorful collage of various abstract shapes and patterns, including circles, squares, triangles, lines, dots, and stripes, representing different artistic styles like cubism or expressionist art. The background is a mix of pink, blue, black, white, and red colors, with elements such as paintbrushes, film reel icons, eyeballs, paint palettes, and computer screens. This design symbolizes the diverse range of visual arts in modern culture. --ar 16:9 --profile i6fyyx5 --v 7
A colorful collage of various abstract shapes and patterns, including circles, squares, triangles, lines, dots, and stripes, representing different artistic styles like cubism or expressionist art. The background is a mix of pink, blue, black, white, and red colors, with elements such as paintbrushes, film reel icons, cats, paint palettes, and computer screens. This design symbolizes the diverse range of visual arts in modern culture. --ar 16:9 --profile i6fyyx5 --v 7
illustration of the bittersweet feeling, abstract, modern, geometric, organic, bold colors, circles, risograph, dots, squares, grids, cat motif, hand-drawn lines, scribbles, textured, distressed, overlay, playful, dynamic, collage, expressive, energetic, overlapping elements --profile i6fyyx5 --v 7
A colorful collage of various abstract shapes and patterns, including an eye in the upper left corner, a film reel at the center bottom with some scattered lines around it, and geometric elements like circles and squares. The background is a mix of pink, blue, black, white, and red colors, with brushstrokes adding texture to each element. In the style of modernism, the design features a flat, vector illustration with bold line work and digital art. --ar 16:9 --profile i6fyyx5 --v 7
illustration of the bittersweet feeling, abstract, modern, geometric, organic, bold colors, circles, risograph, dots, squares, grids, elephant motif, hand-drawn lines, scribbles, textured, distressed, overlay, playful, dynamic, collage, expressive, energetic, overlapping elements --profile i6fyyx5 --v 7
Recent articles
Related article
Free Geeky Animals' Sref code: 4282161627
Important: Remember to use the --sv 6 parameter to quickly determine the version of the Sref code in the future.
Prompts are created using MJ Simple Prompt Generator (+ my brief idea)
. --sref 4282161627 --sv 6 --v 7
Running in circle --ar 1:1 --motion high --video 1
. --ar 1:1 --motion high --video 1
ethereal morning light in a minimalist bedroom, beautiful woman and cat resting together, soft linens, tranquil expression, muted palette --ar 16:9 --sref 4282161627 --profile i6fyyx5 --sv 6 --v 7
Cover prompt: ethereal morning light in a minimalist bedroom, beautiful woman and cat resting together, soft linens, tranquil expression, muted palette --ar 16:9 --sref 4282161627 --profile i6fyyx5 --sv 6 --v 7
I hope you like this article!
Thank you for reading and happy creating!