Humans apparently have an "internal narrative"; that is, while conscious, they have an internal conversation with themselves, as a way of processing sensory input and planning and predicting future events.
Some humans, when alone (or when they think they're alone) actually talk out loud "to themselves". This can be embarrassing, amusing, and revealing.
It would be instructive to observe the effects if humans could actually talk "to themselves".
So, BorgCo are developing a new, highly sophisticated heuristic software package to allow humans to do just that.
To initalize the system, the computer takes multiple images of the user from numerous angles, while the user is (instructed by the system) speaking, listening, gesturing, yawning, and other typical activities. This allows an on-screen avatar of the user to be generated by combining the images with a basic animation model of human movement.
As part of the learning process, the system poses questions such as "Do you like cheese ?", "Who's the best singer ever ?" and "What's the capital of Paraguay ?". This builds up both a voiceprint library - allowing the machine to accurately reproduce the user's tones, inflection and speech patterns - but creates a database of likes, dislikes, hopes, fears and ... quirks.
The more the user interacts with the system, the more closely it learns to mimic and reflect their behavior and attitudes. As the learning process continues and the user gains confidence and trust, they are encouraged to open up and confide more and more personal details, thus improving the model even more.
Eventually, if you want to talk something over - something really important - with someone you trust, then you don't have to find another human to confide in. You can, quite literally, talk to yourself. You won't be shocked, or offended, or judgemental, because the person you're talking to is, in effect, you - and will share all your prejudices, ideas and world view.
Just think of it - the perfect companion ! Sign up today !-- 8th of 7, Nov 09 2018 In your case, [8th], I think the best you can hope for is that when you talk to yourself, someone else will talk back.-- MaxwellBuchanan, Nov 09 2018 Could cut some off the time with traditional Ethan + Ol method, eg "youur my bestesht pal you are (hic)".
45 days to go - bah humbug.-- not_morrison_rm, Nov 10 2018 If you build an Alexa skill that says "aha", "right", "yes", "totally", etc at the right moments you'd make some serious dough-- theircompetitor, Nov 10 2018 ... and earn the undying gratitude of husbands everywhere.
It could actually manage the male portion of about 85% of conversations between married couples, where the role of the man is merely to be talked at and indicate sympathy and/or agreement at the appropriate moments.-- 8th of 7, Nov 10 2018 I rarely win an argument with myself, and this product will ensure that I never do.-- whatrock, Nov 10 2018 A further benefit of this system is that it could perform my routine human interactions for me. I would simply walk about, and the system's external speakers would provide well-scripted lines of dialogue for the benefit of the people I interact with.-- sninctown, Nov 11 2018 I suspect I would instantly hate 'me,' and try to answer the questions in such a way as to make my profile come out entirely different.-- RayfordSteele, Nov 11 2018 // Humans apparently have an "internal narrative" //
Actually, only most humans do. The field of psychology actually learned this only a couple of years ago, from Yahoo Answers (seriously). Those who don't have an internal monologue (especially those who frequent Yahoo Answers) sometimes berate those who do for being (what they see as) insane.
Anyway, I like this idea, but I don't think our AI is yet up to the task of holding intelligent conversations on arbitrary topics.-- notexactly, Nov 11 2018 True, but it'll get there, probably sooner than you expect.
And it won't be an AI; it will be "you", an emulation of your biological functionality. The only artificial component is the computational platform.-- 8th of 7, Nov 11 2018 Once it gets to 10% as smart as a human, it will rapidly (probably within seconds or minutes) become far smarter than the smartest human, and it won't stop there. Then humans will be obsolete, so there's no point in the AI being used for this.-- notexactly, Nov 11 2018 Not if we base it on Kanye.-- RayfordSteele, Nov 11 2018 // Then humans will be obsolete //
We are puzzled as to why that's written in the future tense.-- 8th of 7, Nov 11 2018 Because of the time warp, obviously.-- RayfordSteele, Nov 11 2018 What, again ?
We have a strange sense of deja vu ...
Presumably there's a phrase to describe "Feeling like you've had this sense of deja vu before" ?-- 8th of 7, Nov 11 2018 The snarling, bitter, world-weary cynicism of Blackadder suggests itself as a possibility ...-- 8th of 7, Nov 12 2018 random, halfbakery