Creativity AI #36: Weekly update about AI art
News about AI art, Changes in Geeky Curiosity, Inspiration, Free Style Reference code
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.
Update from Geeky Curiosity
You know why I write Creativity AI newsletter? I want to give you a creative escape from the constant stream of AI updates and daily chaos.
I picture you scrolling through beautiful artwork, experimenting with AI art tools, and actually feeling your creative energy come back. That mental break we all need from the craziness of everyday life.
Starting with the next issue, I'm adding something new. Music! I'll pick a track that matches the vibe of each newsletter perfectly. You can listen while you read. It's completely free. I've been listening to these tracks myself while putting the articles together. They really set the mood.
I genuinely hope you'll enjoy this little addition. There's something special about pairing visuals with sound that just makes everything click.
Alternative link in case the player doesn’t work: —> here
News about AI art
Midjourney just dropped several updates that make video generation more accessible and wallet-friendly. HD video generation is now available for Standard Plan users, though you'll want to use it carefully since it's still expensive.
What's really useful is the new batch size control. You can now generate just 1 or 2 videos at a time using --bs 1 or --bs 2 parameters (or through the menu selection), which cuts your costs by 2x and 4x, respectively. The team still recommends keeping it at 4 for best results, but if you're watching your budget, this gives you options.
They've also separated Moodboards into its own page accessible through the sidebar menu. The varied moodboards create diverse images, while narrow, focused moodboards give you consistent results. It's a great way to combine your real artworks (painting, sketch, illustration) and blend your artistic vision with Midjourney.
Video thumbnails now show the last frame instead of the first, making it easier to tell clips apart. Middle-frame thumbnails are coming soon, which will work better for looped videos. Plus, the moderation accuracy improved, so fewer of your jobs should get incorrectly blocked.
Version 7.1 is in testing with better prompt understanding and improved OREF capabilities. But the real excitement is around V8, which the team describes with cautious optimism as potentially "ridiculously good."
The Style Explorer feature should arrive by month's end, though it might slip slightly. After V8 ships, they'll tackle the V2 video model. The team has an off-site scheduled that might delay releases by a week, and audio features remain on the back burner for now.
Here’s my example of a moodboard art:
woman drinking wine while cuddling a cat --profile ymc22x9 --v 7
ChatGPT-5 rolled out last week with Auto, Fast, Thinking, and Pro modes. That Pro mode needs a hefty $200 monthly subscription, putting it out of reach for most casual users.
For creative work, it shows impressive capabilities. The creative writing feels exceptionally deep emotionally and can mimic specific writing styles convincingly (but I still prefer Claude Opus 4.1).
It is said to be able to handle video production workflows. Feed it one prompt and it generates complete packages: scripts, SEO descriptions, tags, and thumbnail concepts. It can even create functional apps, games, and websites with sophisticated UI/UX understanding. The multimodal collaboration is smooth too, combining text, images, and audio seamlessly.
Based on my limited testing though, Gemini 2.5 Pro still performs better for video analysis tasks. Sora integration isn't native yet, but ChatGPT-5 is said to work alongside it for 20-second videos at 1080p.
I haven't tested all the claims. But I plan to use it more in the coming weeks.
ChatGPT-5 prompt to generate an image:
create a photograph of a cat drinking wine in the bar
Gemini and Claude both got memory features this week, but they work differently. Gemini's approach is continuous learning. Tell it you love sci-fi once, and it naturally incorporates that into future recommendations without you asking. It's rolling out now for Gemini 2.5 Pro in select countries, with plans to expand to Gemini 2.5 Flash soon.
Claude took the privacy-focused route with its August 11 launch. You have to explicitly ask Claude to reference past chats every single time. No automatic profile building happens. It's available for Max, Team, and Enterprise subscribers and can search within specific projects or across all conversations, but only when you request it. Great if you have concern about data privacy and prevent the creepy moment when the bot "knows you too much, too detail".
I tried Wan 2.2 using the free credits, and while text-to-video generation takes its sweet time (you only get one output), the prompt following impressed me. This open-source video generator from Alibaba's Tongyi Lab runs efficiently on consumer hardware, creating 5-second clips at 480-720p resolution.
Got a powerful computer? You can generate videos locally for free. Otherwise, at $6.50-26 monthly, you get cloud access with text-to-video, image-to-video, and editing features. Check it out at wan.video. (Hate the watermark, though)
My prompt using Wan2.2 Plus at 16:9: -
a cat is drinking wine in the bar ElevenLabs just announced the Chroma Awards, which they're calling the "Olympics" of AI Film, Music, and Games. They've partnered with Freepik, FAL, Dreamina, CapCut, and others from across tech, media, and entertainment.
If you're an AI creator, this looks like a solid opportunity. You get to showcase your work, build industry connections, and access free AI tool trials. They're offering $150K in cash prizes plus $1M in free tool trials.
Mark these dates: August 18 for categories and rules announcement, September 1 when submissions open and free trials release, November 3 for submission deadline, and December 7 for winner announcements. Registration is open now, and their Discord has all the updates.
Refreshed Midjourney Style Resources
The Midjourney Style Resources #14 has been updated to V7 aesthetics.
Midjourney Style Resources #14: East Asian Aesthetics
This article is part of the resource series "Geeky Curiosity's Boutique Collection of Midjourney Styles".
Inspiration
This week I'm captivated by movements: how static images can trick our minds into feeling things are moving. Not the obvious motion in AI videos, but those subtle visual cues that make our hearts race when looking at a still frame.
bride standing on cliff’s edge, wind whipping her veil in dramatic arcs, ocean waves crashing below, overcast lighting, serene and sorrowful emotional contrast --ar 16:9 --profile 3cavmty tt3bo9d --v 7
emotional goodbye hug at a train station, blurred motion of departing train, tears streaming, heavy contrast lighting, real raw emotion in the grip and expressions --ar 16:9 --profile 3cavmty tt3bo9d --v 7
child sprinting through a sunlit meadow, arms wide open, golden hour light streaming through tall grass, soft motion blur on background, expression of pure freedom and joy --ar 16:9 --profile 3cavmty tt3bo9d --v 7 
person running through a rainstorm, drenched and illuminated by streetlights, puddles splash dramatically, captured in motion blur, melancholic and intense emotion in the moment --ar 16:9 --profile 3cavmty tt3bo9d --v 7 
a ballerina mid-leap, body arched in perfect form, dramatic stage lighting casts sharp shadows, motion captured at apex, emotional intensity in posture and face, darkened background --ar 16:9 --profile 3cavmty tt3bo9d --v 7 
mother spinning child joyfully on a beach, blurred motion and sand flying, waves crashing, sunset casting long shadows, emotional warmth and connection --ar 16:9 --profile tt3bo9d --v 7 
Recent articles
Related articles
Free Geeky Animals' Sref code: 4132772101
Sref code version: --sv 4
Prompts are created using MJ Simple Prompt Generator (+ my brief idea)
Bustling open-air market scene, colorful produce stalls, people smiling and chatting, vibrant tones, warm daylight, casual joyful vibe --ar 16:9 --sref 4132772101 --profile tt3bo9d --sv 4 --v 7 
illustration of neighborhood block party at dusk, dancing under string lights, food trucks, joyful crowd interactions, warm tones --ar 16:9 --sref 4132772101 --profile tt3bo9d --sv 4 --v 7 
illustration of a busy Cozy coffee shop interior, people laughing, working on laptops, soft lighting, rustic textures, friendly and relaxed ambiance --ar 16:9 --sref 4132772101 --profile tt3bo9d --sv 4 --v 7 
Cover prompt: illustration of neighborhood block party at dusk, dancing under string lights, food trucks, joyful crowd interactions, warm tones --ar 16:9 --sref 4132772101 --profile tt3bo9d --sv 4 --v 7
I hope you like this article!
Thank you for reading and happy creating!