The dominant AI paradigm focuses mostly on increasing compute and data to improve model capabilities. While these models will continue to advance, they remain "jagged" and dramatically inefficient compared to the human brain. Given that we have living proof of a more sophisticated, high-efficiency path to intelligence, exploring alternative architectures is a strategic imperative. Astera offers a radical departure from the “scale is all you need” philosophy, instead turning to neuroscience to decode the biological blueprints of the mind. By investigating how natural systems achieve profound cognition on the power consumption of a lightbulb, Astera is prioritizing the discovery of fundamental algorithmic principles that might lead to the next generation of efficient, elegant intelligence. 

Astera enables the Neuro-AI division to focus on basic research and take a long-term view, while operating nimbly like a startup. We work in teams on difficult milestone problems unconstrained by the need to secure funding, garner profit, or publish results. By providing significant resources, Astera provides a home to diverse AI research that would otherwise be neglected. The team also has access to significant computational resources.