Responsible AI - Conceptual Search and Digital Brains

 




Our Responsible AI efforts tie in with our research on advanced architectures for conceptual search and cognition. 

Other companies are trying to solve these problem using abstract tensor models.

Noonean uses a more brain like architecture representing language relationships. This next architecture target fits better into a responsible AI target because we can better develop limits and rules for it as it is not a mathematical abstraction.  So while companies like OpenAI are funded with billions of dollars they produce little related to responsible AI. 

Our longer term goal is to develop from conceptual search engines like SULU (our conceptual search product) to digital brains (Noonean Cognitive Engine) used for autonomous systems like androids, and also for more intelligent knowledge bases. We don't believe two dimensional Tensor or Transformer architecture will get there.  These architectures are best for image recognition and finite state machines but not for cognition or language.  We will be uniquely positioned to lead in this growing market for autonomous systems providing the digital brains in the same way Intel today provides the CPU for a computer. 

What Makes the Noonean Approach Better?

We utilize a very different architecture based on three dimensional neural units which are dynamically relating and organizing, rather than fixed 2D planar models with tensor calculations between planes. Systems that focus on language and awareness, behaviour and communication will never become very advanced using the Hinton two dimensional planar neural network approach. 

So while other companies are working on restricting non-woke language, we are actually evaluating language based intentional and cognition systems and following I Robot models of agency.