Safe Superintelligence has secured a deal with NVIDIA for priority access to its upcoming Vera Rubin chips. This comes as global AI investment surged to $243.9 billion in 2025, per PitchBook.
TL;DR
- Safe Superintelligence gains exclusive access to NVIDIA's next-gen chips, accelerating its superintelligence research.
- Global AI investment hits $243.9 billion in 2025, with major labs driving progress toward artificial general intelligence.
- AMI Labs focuses on world models, while Safe Superintelligence prioritizes safe superintelligence development.
What happened
Global AI investment reached $243.9 billion in 2025, accounting for more than half of all venture capital deployed worldwide, according to PitchBook. This surge in funding has accelerated the development of artificial general intelligence (AGI) across major international research hubs.
Safe Superintelligence, a lab with approximately 50 employees, has secured a deal with NVIDIA for priority access to its upcoming Vera Rubin chips. This agreement will support Safe Superintelligence's training clusters and accelerate its research towards safe superintelligence.
AMI Labs, another key player in the AGI race, is focusing on world models that construct internal representations of physical reality. This approach differs from traditional sentence prediction methods. Co-founder David Silver emphasized the importance of AI learning from its own experience rather than human-generated datasets in a research paper detailing the operational thesis behind Ineffable Intelligence.
Why it matters
The substantial increase in AI investment highlights the growing emphasis on AGI development. This trend presents opportunities for startups and investors to participate in the rapidly evolving AI landscape.
Safe Superintelligence's deal with NVIDIA underscores the importance of specialized hardware in AGI research. This agreement could give Safe Superintelligence a competitive edge in developing safe superintelligence.
AMI Labs' focus on world models represents a distinct approach to AGI development. This could lead to advancements in AI's ability to understand and interact with the physical world, opening up new possibilities for AI applications.
Key facts
- Global AI investment reached $243.9 billion in 2025, per PitchBook.
- Safe Superintelligence has secured a deal with NVIDIA for priority access to Vera Rubin chips.
- Safe Superintelligence has approximately 50 employees.
- AMI Labs is focusing on world models for AGI development.
- David Silver is the co-founder of AMI Labs.
- Ilya Sutskever is the founder of Safe Superintelligence.
- NVIDIA's Vera Rubin chips will support Safe Superintelligence's training clusters.
Context
The race towards AGI is intensifying, with major labs like Safe Superintelligence and AMI Labs leading the charge. These labs are exploring different approaches to AGI development, from world models to safe superintelligence.
The substantial increase in AI investment reflects the growing recognition of AI's potential to transform industries and societies. This trend is expected to continue, with more startups and investors entering the AI space.
The competition for specialized hardware highlights the importance of infrastructure in AGI research. As AI models become more complex, the demand for powerful and efficient hardware will continue to grow.
