On thinking about what would be the creative usage, I started with what kinds of input a person can give.

• Text
• Body movement
• Interaction with physical device
• ETC (Showing image to something, holding some objects that a system would like)

I found body movement interesting, especially when its absence is used as input, not its existence.

I started thinking about a scenario where computers(AIs) want to talk to each other like dogs want to know other dogs. As these devices are here to help people and work under the control of people, I thought these questioning will happen when the owner is not around.

Under freedom of supervision, the mac tries to call the peripheral device, iPhone, to ask a bunch of questions until the point it bothers it. Humans find that something weird is happening to iPhone's Siri until one finds out that it was MacBook that was continuously calling iPhone for its response.

Please increase the speaker volume when watching the video

Please increase the speaker volume when watching the video

Code

I used a simple PostNet example from Dan to recognize whether a person is near or in front of the computer.

I changed the voice to that of Samantha as it was a voice that Siri was able to hear most clearly among different options of voices.

Problem unsolved

I tried using the following code**(cancel())** to stop when the user is detected and also tried putting break time between Macbook's one question and the other questions using setTimeout.

However, after I combine these two and human got detected, it started to make noises like "Siri, si, si, si, si" as if there are arrays of things to say but has to cancel one by one until they exhausted all the arrays.

synth.cancel();

setTimeout(function(){
      console.log("Human gone"+num);
      speak("Hey siri, " + "......." + siriQuestions[num].text);   // speak the text inside the ()
    }, 10000);