![Cognitive Architecture for GPAI](/images/2023/07/26/blog-img-23_large.jpg)
Perceptible builds cognitive models for GPAI in the real world. Using psychology and transactional analysis, Perceptible builds cognitive models that interact with people in a way that is safe and efficient. TA-based autonomous agents are made to think and act as people,
and interact in a way that is acceptable in a future where autonomous agents effectively live and work among us.
Founded on principles of psychology and transactional analysis, Perceptible builds AIs that think and act like people. This human-centric approach allows autonomous agents to understand social dynamics and have natural conversations. Perceptible it equips autonomous agents with emotional intelligence so they can empathize with users and adjust their behaviour accordingly.
We aim to make AI more humanistic by giving it emotional intelligence and social awareness. Our agents can communicate with nuance, adjust behaviors to avoid harm, and cooperate with people. We simply believe this approach results in AI that is more useful, safe and comprehensible.
For example, AI assistants built using Perceptible's approach can adeptly handle conversations and tasks without needing extensive programming. Our goal is to equip service robots with social awareness to seamlessly integrate into human spaces.
Central to this vision is to approximate human behaviour that is expectant of quality collaborators within the cognitive models we produce.
By adopting a humanistic and psychology-first approach to AGI we proactively position our cognitive models to be socially aware.