Wait, what? You can now create open interactive worlds from a single image
All the news from the AI world for creatives
Good day, creatives
In today’s Newsletter:
1️⃣ Wait, what? You can now create open interactive worlds from a single image
2️⃣ Amazon joins the AI party with 6 new models
3️⃣ Pentagram used Midjourney for a project. Learn how.
4️⃣ Quick news you may have missed
5️⃣ Prompt Fight
6️⃣ Video of the Week
7️⃣ Steal this Midjourney sref
Let’s go.
Wait, what? You can now create open interactive worlds from a single image
Two new announcements related to AI-powered 3D world generators were unveiled this past week: These tools can transform screenshots into interactive, navigable 3D environments.
Read that again.
This technology could redefine industries like gaming, film, and simulation by making world-building faster and more immersive.
Here are the news:
First, Stanford professor Fei-Fei Li, dubbed the "godmother" of AI, released an early peek at what her AI startup has been working on: an engine that can turn still images into realistic three-dimensional worlds.
Li's World Labs demonstrated its approach, sharing examples of realistic photos and fantasy images being turned into 3D worlds, including one based on a Van Gogh painting.
Then, Genie 2
Just days after Fei-Fei Li’s announcement, Alphabet answered with its own interactive 3D platform Genie 2. It can generate interactive digital worlds based on screenshots and text alone:
This technology could redefine industries like gaming, film, and simulation by making world-building faster and more immersive.
Why this is important
The creative potential unlocked by these tools is becoming clearer. Whether in training AI, creating digital art, or revolutionizing storytelling, tools like Genie 2 and World Labs’ engine are reshaping what’s possible in both virtual and physical worlds.
Amazon joins the AI party with 6 new models
Amazon unveils model family, they showed off not one, not two, but six new Nova models:
Amazon Nova Micro, a text-only model that delivers responses at very low cost.
Amazon Nova Lite, a very low-cost multimodal model that is lightning fast for processing image, video, and text inputs.
Amazon Nova Pro, a highly capable multimodal model with the best combination of accuracy, speed, and cost for a wide range of tasks.
Amazon Nova Premier, the most capable of Amazon’s multimodal models for complex reasoning tasks and for use as the best teacher for distilling custom models (still not available).
Amazon Nova Canvas, an art image generation model.
Amazon Nova Reel, a video generation model. (Examples here)
Why this is important
It seems the party is almost complete, now. Every tech giant (but Apple) has made its move to offer native AI models.
Pentagram used Midjourney for a project. Learn how.
A 50 year old design firm, @Pentagram, known for its craft (typefaces), used Midjourney to create illustrations for a recent brand project. And people were not happy. But they shared their workflows for the illustrative style, and they are interesting.
Quick news you may have missed
KREA AI released a new AI trainer, allowing you to train your very own AI image generator model.
We have a new Chinese AI Video tool: HunyuanVideo from Tencent is a 13B parameter open-source(!) text-to-video and video-to-video model focused on high quality video generation with physical accuracy and scene consistency.
Anthropic’s Claude can now clone your writing in 1 click, it offers style customization, with options like formal, concise, and explanatory, or the ability to mimic your personal style using writing samples.
Prompt Fight: Pixels
Lets put the best AI Art Generators to the test with the same prompt:
Detailed rendering of a 10000000 pieces of small cube-shaped pixels in a voxel-based 3D illustration depicting an isometric cube made from apple, set against a white background, in an 8-bit, minimalistic design, with a front view, created using Octane rendering with a 256-color palette.
AI Video of the Week
Check out The Batman, an impressive 10-minute film made by Kavan Cardoza.
Steal this Midjourney SREF Code
To use it, add --sref 3486083104. at the end of your prompt in Midjourney. Your images will be generated with this style.
Read more: Power up with these 10 Midjourney Sref codes for your AI Art