TL;DR:
- Stability AI introduces Stable 3D, an AI-powered 3D model generation tool for artists and developers.
- Stable 3D aims to simplify 3D content creation, making it accessible to non-experts.
- Users can describe or upload images for conversion into 3D models, enhancing creative flexibility.
- Uncertainty surrounds the training data source, raising potential copyright concerns.
- The competitive landscape includes established players like Autodesk and newcomers such as Kaedim and Luma.
- Despite financial challenges, Stability AI recently secured $25 million in funding.
- The company is diversifying with new features in its AI-powered photo editing suite.
Main AI News:
Stability AI, the trailblazing startup renowned for its text-to-image AI model Stable Diffusion, is pioneering the future of generative AI with its latest offering. Enter Stable 3D, a game-changing AI-powered application designed to revolutionize 3D model creation for platforms like Blender, Maya, Unreal Engine, and Unity.
Accessible exclusively through private preview for select customers who engage with Stability via the company’s contact form, Stable 3D is set to empower even non-experts to produce “draft-quality” 3D models in mere minutes, according to Stability AI’s recent blog post.
“For graphic designers, digital artists, and game developers, 3D content creation has traditionally been a complex and time-consuming endeavor, often taking hours or even days to craft a moderately intricate 3D object,” the company states. “Stable 3D levels the playing field for independent designers, artists, and developers, enabling them to churn out thousands of 3D objects per day at minimal cost.“
While it may seem like lofty rhetoric, Stable 3D appears to possess robust capabilities, comparable to other model-generating tools on the market. Users have the flexibility to describe their desired 3D model in natural language or upload existing images or illustrations for conversion into 3D models. The output is delivered in the “.obj” file format, which allows for seamless editing and manipulation using standard 3D modeling tools.
One crucial aspect that remains undisclosed is the dataset used to train Stable 3D. Given the proclivity of generative AI models to replicate training data, this could potentially raise concerns among commercial users. If any of the data used for training is subject to copyright restrictions and Stability AI lacks proper licensing, Stable 3D users could inadvertently incorporate IP-infringing content into their projects.
Stability AI has previously faced allegations of IP infringement when Getty and several artists sued the company for allegedly utilizing their images to train Stable Diffusion without proper notification or compensation.
Although Stability AI recently partnered with startup Spawning to address “opt-out” requests from artists, it remains unclear whether this partnership covers the training data for Stable 3D. We have reached out to Stability AI for further clarification and will update this article accordingly.
Legal considerations aside, Stable 3D signifies Stability AI’s foray into the burgeoning but competitive realm of AI-powered 3D model generation. The landscape includes platforms like 3DFY and Scenario, along with startups such as Kaedim, Auctoria, Mirage, Luma, and Hypothetic. Even established players like Autodesk and Nvidia are exploring this space with offerings like Get3D, which transforms images into 3D models, and ClipForge, which generates models from textual descriptions.
Additionally, tech giants like Meta and OpenAI have ventured into 3D asset generation. OpenAI, in particular, released Point-E, an AI capable of synthesizing 3D models with applications in 3D printing, game design, and animation.
Stable 3D stands as Stability AI’s latest strategic move to diversify and adapt in the face of mounting competition from AI platforms specializing in art generation, including Midjourney and the aforementioned OpenAI.
In April, Semafor reported that Stability AI was facing financial challenges, prompting an executive hunt to bolster sales efforts. Forbes reported instances of wage and payroll tax delays or non-payment, leading AWS — the provider of compute resources for Stability’s model training — to threaten the revocation of Stability’s GPU instance access.
Despite these challenges, Stability AI recently secured $25 million through a convertible note, augmenting its financial reserves to over $125 million. However, the company has not yet secured new funding at a higher valuation, with its last valuation standing at $1 billion. Stability has expressed ambitions to reach a valuation four times higher in the coming months, despite persistently low revenues.
In a bid to differentiate itself and stimulate sales, Stability AI has unveiled new features for its AI-powered online photo editing suite. These enhancements include a model fine-tuning feature that allows users to personalize the underlying art-generating models, as well as a “sky replacer” tool that seamlessly substitutes the color and aesthetic of the sky in photos with preset alternatives.
These new tools join Stability AI’s expanding portfolio of AI-powered products, which already includes the music-generating suite Stable Audio, the doodle-creating app Stable Doodle, and a ChatGPT-like chatbot.
Conclusion:
Stability AI’s Stable 3D marks a significant step in democratizing 3D content creation, potentially leveling the playing field for artists and developers. However, concerns about data sourcing and financial stability remain. The 3D model generation market is intensifying, with established players and startups vying for a slice of the pie, setting the stage for increased innovation and competition.
