We wanted to set up a dialogue at the heart of our game to allow people to have personalized advice and support in their pocket available at any time. Indeed, there are many applications to stop smoking, usually filled with statistics, graphics, which we believe lack in human contact. Ikta also allows, by its design, to put a little distance between the possible judgments that the player could otherwise suffer from their family, their friends or society.
We decided to use the VIDE asset to develop our dialogue since our application is developed directly in the Unity game engine. VIDE (which has a free version) is very easy to use and allows for data recovery via the dialogue between Ikta and the player (name, concentration of nicotine, etc.). VIDE does not offer artificial intelligence to manage the dialogues but rather a tree where the path depends on the player’s choices. Due to the lack of space on the screen, we can only leave two choices to the player. This prompted us to create an introductory series of dialogues to better under the player and their desire to smoke. After a few days of use, we started to write a series of dialogues on a wide range of topics such as getting over a craving, defining one’s reduction goal, relaxation tips, etc. These dialogues do not have to be presented in any specific order.
In the future, we imagine the dialogues to be a full chatbot rather than just a pre-programmed skit. Ideally, the user will be able to talk with Ikta about whatever they want using their own words. This would provide a feeling of closer connection between the player and Ikta, and would result in a more personalized dialogue that is relevant to the player’s needs. To give an example of the kind of interaction the player could have with Ikta, we could imagine a scenario where the player has an intense craving on a certain day. They would be able to connect to Ikta and specifically ask her in their own words “Can you give me some tips to get over a craving?” Ikta would then provide some techniques or links to relevant websites that the user can consult at their own leisure. The major challenge of such a chatbot would simply be integrating it into the C# code.
It is clear that the content in these dialogues needs to be verified (and even directed!) by medical and psychological experts to avoid ineffective (or worse, counterproductive) content written by game programmers with good intentions but who don’t really know what they’re doing.