Since the end of last semester, I started to have this self church idea of how most time we are imprisoned by our own logic and brainwashed by ourselves. One of the articles I read discussed about how self-criticism becoming self-hypnosis:
Self-criticism, when it isn’t useful in the way any self-correcting approach can be, is self-hypnosis. It is judgement as spell, or curse, not as conversation; it is an order, not a negotiation; it is dogma, not overinterpretation.
And meanwhile, what happened in the past few month made me even more strongly feel that conversations most times are encoding and decoding. All the communications are based on our experiences and how we think. We only hear what we want to or are able to hear.
Therefore I want to have this avatar containing multiple “selves” of me. When you speak to them, the program analyzes your words and only looking for certain keywords. If there’s any sound in your words found similar to these keywords, the “controller” of “selves” will be triggered, a hypnotic voice will speak the corresponding sentence and then “selves” will have to follow. Only when the person make really loud noise, “selves” will be released and be themselves. So this is an avatar that you could control it but it still only think and behave like how I think. The build-in sentences are the sentences I’ve been always thinking or trying to convince myself.
Now I have this processing sketch just randomly picking build-in sentences and all “me”s follow. When it detect very strong noise, “me”s will do different things. But if the noise stop for 10s, the cold voice will control me again.
For next step, I need to increase the behave library, and then find a way to turn speech into text. There’s a cloud speech api which I use but I need to put everything into web base, or use processing library .