One of Europe’s top AI researchers raised a $13M seed to crack the ‘holy grail’ of models

SpAItial, a startup founded by Matthias Niessner, aims to develop AI foundation models capable of generating full, coherent 3D online environments from text prompts. Niessner, a prominent AI 3D model researcher, has raised $13 million in seed funding led by Earlybird Venture Capital. The startup's technical team includes experts like Ricardo Martin-Brualla and David Novotny, who bring significant experience from Google and Meta, respectively. SpAItial's goal is to create interactive 3D spaces that behave like the real world, with potential applications in video games, entertainment, construction, and robotic training. The company plans to license its foundation model to developers for specific use cases and is focused on generating revenue early on.

Despite competition from startups like Odyssey and World Labs, SpAItial is working on creating larger and more interactive 3D environments. Niessner envisions a future where users can easily create their own video games, bypassing the restrictions imposed by current gaming platforms. The startup is prioritizing quality over quantity in its team and aims to partner with developers to explore the potential of its models. SpAItial's ambitious vision includes replacing traditional CAD systems, marking the beginning of a new chapter in 3D generation.

Key takeaways

SpAItial, founded by Matthias Niessner, aims to create 3D online environments from text prompts, similar to how AI models generate images.
The startup has raised $13 million in seed funding, led by Earlybird Venture Capital, despite having only a teaser video to show.
SpAItial's team includes experts like Ricardo Martin-Brualla and David Novotny, who bring significant experience from Google and Meta.
The company plans to focus on generating interactive 3D spaces and aims to license its foundation model to developers for various applications.

One of Europe’s top AI researchers raised a $13M seed to crack the ‘holy grail’ of models | TechCrunch

Key takeaways

Discussion (0)