Do you remember the Akinator back in 2007? Akinator is the online genie who would guess any fictional or non-fictional character in your thoughts within an average of 20 questions. I spent hours certain I could defeat the Akinator by thinking of a character it would never be able to guess. I never suceeeded.
Even though the Akinator launched more than a decado ago, and still remains popular today, its level of sophistication is higher than you’d expect if compared to other bots in the market.
Thanks to the Maturity Model above, I can assess its current level of maturity and there are clear examples on how the Akinator could be taken to the next level on all
Akinator is a multi-channel game which can be played on an app, Facebook and website, available in over 10 languages. The same experience is also available through a voice interaction on Alexa. Furthermore, the bot completes the entire interaction without the assistance of a human to provide the responses.
In order to be considered at the next level of maturity, Akinator could be an experience which begins on one channel, and can be continued on another.
There are multiple theories on how the Akinator collects this incredible amount of knowledge. The most plausible one seems that it uses Machine Learning to know which questions to ask and scraps the web for answers. But this information has been kept secret from the public. It is also impressive that the Akinator provides the answers directly within the bot interface and does not redirect the user externally to retrieve the final answer.
In order to be considered at the next level of maturity, as the Akinator collects key insights on the user’s favourite characters, it could diversify its content by providing complementary information – for example, a movie recommendation if the character is an actor/actress. Furthermore, the bot could initiate the conversation with the user when detecting diverse factors, such as boredom for example.
Akinator provides a set list of potential responses for the user to choose from: Yes, Don’t Know, No, Probably, Probably Not. The current responses are static, and lack a personal touch.
In order to be considered at the next level of maturity, Akinator’s interaction could be more detailed; it could remember facts about the user’s interests and mention them in order to make the conversation more personal.