NeuroSynthetica is building the foundation for the next generation
of Artificial Intelligence-- Synthetic Intelligence. Unlike the rule-based
expert systems of the 1970s and Machine Learning programming paradigms of 2010s,
Synthetic Intelligence is artificial intelligence through sentience, arising
from a simulated synthetic model of the minimal brain architecture necessary and
sufficient for Behavior to emerge.
In collaboration with our partners, we are isolating that architecture and
enabling its deployment in commercial applications.
NeuroSynthetica is committed to solving sentience and enabling its safe and
effective deployment in synthetic form.
Our scientific inquiry seeks to understand the conditions under which sentience
arises, in natural biology as well as in simulo. Specifically:
- How does Sentience
arise in brains, natural or synthetic?
- What is the minimal brain
required for sentience to arise?
- A minimal brain architecture exists, which is necessary and
sufficient for sentience to arise.
- The emergence of sentience is invariant with respect to the
underlying host substrate (in vivo, in silico,
or in simulo).
- The sentience effect is observable in a simulated model's
as it interacts with the external world through
- Working with biological and computational biology researchers, we help
build models of candidate architectures to test implementation hypotheses.
- The experimental results are used to refine models and support further
inquiry, and may be made available as technology which can be used to
solve commercial AGI problems.
To make our inquiry possible, we've built simulation server technology that
simulates at-scale brain models (>1bn neurons with 1ms temporal
resolution) created with graphical tools using
SOMA™, a descriptive modeling
language. The simulation server and model development tools run on
commercially-available X64 PC and server hardware.
To allow models to interface with the physical world (or a virtual one such as the
webscape on the internet), we have created robot interface libraries, which are
available for use by community developers using platforms ranging from Raspberry Pi
to X64 PC systems.
All our software runs on Linux.
Our commercial business involves the licensing of technology and products such as the
NeuroSynthetica Sentience Engine™
to commercial and US government customers. While we are investigating sentience,
our early-adopter commercial customers can get ahead of the game and use our commercial
simulation engine to power real-time Machine Learning (ML) models using next generation
In our scientific role, we support the brain research community by offering the use of
our technology and products, and enabling collaborative information sharing, through
our Subscription Service. Subscribers
can download the NeuroSynthetica Sentience Engine simulator, the Workbench development
environment, robotic interface libraries and sample robot code. Documentation is