Kinesthetic experience and machine learning
New topic and new material, and I think both are interesting. I haven’t worked with machine learning before. I just read about it during my AI summer course and I am looking forward to getting some experience of it. I think it is interesting the concept of learning and what it means in the context of machine learning vs human learning. How machine learning and its intelligence is a representation of what it has been trained on. I think this is interesting since there is more to it than just if it recognizes a banana as a banana when it comes to what data the machine has been trained on it also affects how inclusive the design will be. In our case, we will work with detecting body parts. One thing I wonder about is what kind of body has it been trained on and how will it interpret deviations. If I understood how the program we are going to use works correctly then it wouldn’t understand that a person can have let’s say only one leg. It will predict where the other legs knee, ankle would be. And this doesn’t feel quite right and in some cases, there even might be a risk for excluding persons that are outside of the data that the machine learning has been trained on. Design ethics and accessibility are something I wish we would work more with during the education because I would like to learn more about how to design systems that value accessibility.
How we will work with the kinesthetic and use that to design a kinaesthetic interaction still feels a little bit unclear. What are we trying to design and what will the output from a movement as an input be? Sometimes it feels that we are a bit too free to choose our own exploration and it would be nice to get a little bit of direction.