hero_image
Our Approach to AI
Leveraging a creative sidekick
AI is here to help
Let LISA do the heavy lifting
Our aim is to provide clients with powerful assistance to improve their working lives, the quality of their exercises and the capability of their teams.
AI won’t replace you
We view AI as a fast, creative sidekick to whom you can delegate tasks but not responsibility. LLM AI like OpenAI are statistical models that seem intelligent but work on the basis of probability, not reasoning. This means that while they usually give incredible results, they can hallucinate and get things wrong.
Verifiable results
We aim to make all AI-generated results verifiable by you, the real expert. This means when the AI is asked to score a player response, it explains the basis for the score and allows you to modify it. Just like a GPS navigation tool like Google Maps, it can’t see the current real world circumstances like road closures or broken traffic lights. Never be afraid to doubt the AI’s answers.
Human-like
Realism is important for engagement and AI can help create content and interactions that mirror humans. Our goal is to make persona reactions and the content they generate more human-like but allow them to have opinions and points-of-view (i.e. beliefs).
A model displaying the Conducttr crisis simulation approach to employing AI in crisis exercises.
Conducttr’s AI Model
The image on the left represents our model for how AI is employed in creating and delivering realistic exercises.

The World represents the physical and information environment in which exercises take place.

The Scenario is the focus of the exercise.

Domain Expertise is knowledge applied and embedded by Conducttr
Tthe domain expertise of Conducttr’s AI in creating realistic, engaging crisis exercises and crisis simulations.
Domain Expertise
Conducttr’s AI incorporates over a decade of experience in participatory storytelling to ensure content and behaviours are realistic and engaging.
AI In Information Operations Exercises
Download our whitepaper on AI in information operations exercises prepared for the ITEC conference

Download