Natural Learning

Through interactions with people, our technology builds a shared understanding of the human experience so that it can become a fluent thought partner and facilitate complex decision making. Our hope is that the Elemental Cognition technology will be able to learn to answer and explain well enough for humans to take individual responsibility for any decision made on the basis of those answers.

At Elemental Cognition, we are teaching our AI system to learn in ways similar to how children learn: through reading, collaborative dialog, and reasoning. Our system leverages statistical machine learning while focusing on what we call natural learning, enabling humans and machines to learn from one another, share understanding, and develop rich explanations together.

 
 
 
 

How does Elemental Cognition work?

 

From The Elemental to The Complex

Our system learns to understand human language like we do: it gets better at understanding more complex ideas by learning and building on simpler ones first.

 

Layered Learning

The system interprets language at multiple layers of meaning (i.e. textual, grammatical, semantic, logical, and episodic) and then builds a causal model of understanding, capturing how and why things make sense while identifying gaps and opportunities to refine and expand its understanding. 

 

Continuous Human-Machine Collaboration

The system takes steps to build a richer understanding through continuous dialog with people and other sources of incremental knowledge. As the Elemental Cognition system collaborates with people and digs deeper, it builds an increasingly sophisticated understanding, learns to formulate smarter questions, and provides deeper explanations. This is what we call natural learning.

 

Creative Potential, Amplified

Over time the system becomes more effective, accelerating our ability to reason and helping us solve more and more challenging problems. Partnering with this kind of AI will result in amplifying a person’s ability to solve more problems and create better solutions. 

 

Artificial Intelligence through Human Cognition

Our architecture is shaped by how people think and learn: the shared experience of our world is the primary input and language is the secondary input. Elemental Cognition’s technology strives to build a shared understanding of the human experience through dialog wth people and grasp how humans think. In that way, the elements of human cognition are the foundation of our technology. They are the reason we can provide meaningful explanations to people when they ask the question, 'Why does this answer make sense?’