Two years ago I have done a brief introduction on Artificial Intelligence for those who just finished high school and want to get in universities (which I wrote about before, here.)

If I would do it again now, I’ll do it completely different. My grasp of AI technologies and how much they emerged in our life has been widened upon the narrow view of regular AI approaches. From Google glass and wearables to Xbox One and leap motion, the hardware can deliver awesome platforms for us to implement amazing ideas into our everyday life. From start-ups and one-man-team to indie developers, we are not limited, anymore, with resources and old-fashion ideas (I think the ideas of AI presented in the slides are pretty much old-fashioned and they represent what I was learning in university and they don’t represent the real potential of AI.) For instance, my work on games for the last two years really opened up my mind on many issues we will face in the upcoming years. I worked on emotions modelling on two separate projects (in 2012 and 2013, see the publication section for this paper: “A Quantitative Approach for Modeling and Personalizing Player Experience in First-Person Shooter Games” and my upcoming work for UMAP conference 2014 “Towards More Accurate Player Experience Models: An Exploration of Context, Behavioral, Visual and Affect Features”.)


My fourth year project in 2012, and thanks to Noor Shaker who proposed and supervised the implementation of the idea, was about personalizing content generation in first person shooter games through player modelling (don’t be afraid from the hilarious name) in a complete state of the art approach (here for more). I do admit that, upon finishing the project successfully, I was amazed of what our game can do using the model we built. The model can intelligently manipulate the emotions of the player by automatically generating specific game content tailored by the player actions and hisher style while playing the game. The model can for example auto-generate a game level that can maximize the player’s engagement in the game or even maximize the player’s frustration! This is a two-edged sword since this can be implemented to obtain certain goals to the authoring party or used in a wrong way.


What I’ve been working on now (2013 and ongoing and supervised by Noor Shaker and Julian Togelius from ITU, Denmark) is an authoring tool for Cut the Rope, named Ropossum V1.0. My own version of the game is named Cut the Rope: Play Forever. Ropossum is all about letting you design your own levels, check your designed levels for playability on real time. You can also ask it for help to complete your unfinished designs according to your own preferences or just play the game you want forever only by asking Ropossum to generate a never-ending levels for you. A more futuristic vision is to create a social community for gamers, letting the players rate the designed/authored/generated levels and share them with their friends to create a whole new social community for games generating themselves by the effort of the players!

What we can see is that the games now still fall short on technologies that are proposed a while back. We should enjoy the technology we have because up to now, we are, sadly, lacking the courage to go through into the next phase.

To be continued!

One response to “Artificial Intelligence! Now what?”

  1. ArtificiaI Intelligence brief Introduction @Wikilogia – Freshman 2012 | Mohammad Shaker Avatar

    […] Update: for an update and a self-discussion followed by this post go to this post. […]


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at

%d bloggers like this: