OS-tan theories and disscusion

[Smokey]

Maybe a couple of rulers... . You could end up with something nice there...

[NejinOniwa]

She would be the RULER of rulers!!! XD

[Bella]

Hey, don't blame me, I like the metric system. It makes one helluva lot more sense than the US system; but I don't use it much outside of school : \

Metric-tan would be pretty neat, though...

Are we getting dangerously OT? Nobody even commented on Solaris-tan and her incredible magnetic field

[NejinOniwa]

I know, I know...but I'm too busy trying to multitask between maths work and forum posting so I can't help much >_> Sori! ;

[Toki]

Well, the topic is OS-tan Theories... Ehhh I don't know anything about Solaris-tan I'm sorry ;-; I'll draw her for you though if you give me a reference/description...(?)

[Bella]

I know, I know...but I'm too busy trying to multitask between maths work and forum posting so I can't help much >_> Sori! ;

Maths, huh? I sympathize with you, brother! XD

Well, the topic is OS-tan Theories... Ehhh I don't know anything about Solaris-tan I'm sorry ;-; I'll draw her for you though if you give me a reference/description...(?)

Thanks for the offer, Toki! Solaris-tan has been drawn before by Aurora (of course, this is a pretty early drawing of hers'):



And I've drawn her before in group pix (she's the redhead on the far right):



But physically, it's fairly agreed she has medium-short choppy orange hair, purple-blue eyes, and is of a rather normal build. She sometimes dresses like a waitress, but also enjoys dressing up quite elaborately, with lots gold and sun themes in her clothing. She really loves coffee, too; her special Java to be exact. While a very smart and competent, she has a very quirky side, too, much of which revolves around a strange admiration/rivalry with Linux-sama.

I could babble on, though; drawing her is up to you, but I think it'd be neat to see a Toki rendition of Solaris-tan

[Toki]

Hm! I'll see what I can do! I do have a soft spot for redheads.

[Smokey]

I know, I know...but I'm too busy trying to multitask between maths work and forum posting so I can't help much >_> Sori! ;

Maths, huh? I sympathize with you, brother! XD

Peh, Maths is eeaasyyy... it just bores the crap out of me, unless it involves some interesting "stuff" ;010

[SCAI]

A few days of research in preparation makes every day in the field easier.


But to bring a new theory to light, as is the topic for this thread...
If developed as a system, how would the OS-TAN function?
After running into this place my interest for Artificial Intelligence development was once more piqued. Seeing the level at which independent/complete AI's today are - meaningly, solely theoretical and primitive - this might feel like rushing things a bit, but I normally prefer to plan ahead quite far; especially with sensitive subjects as this.
So to wrap things up, just how would a real OS-TAN function, in collaboration with the User and the rest of the system?

[Smokey]

good question...

And as for the AI, I'm trying to figure out a basic learning protocol...But it's not as simple as i thought to implement...

[SCAI]

Verily so. Learning protocols can be anything from simple to complex in theory, but they're never easy to implement well.
But on the subject of OS-TAN, the outcome is meant to have a human disposition and mentality, correct? Then most likely a Turing-complete simulation would be the ideal type - instead of having to implement all kinds of complex algorithms and get a software piece of unknown capabilities.
Another theory that an acquaintance of mine formulated was this:
Two functions are necessary. 1: A function that relays a GOOD or BAD signal to the program, similar to the human system of pain and pleasure. 2: A function to write results from previous actions and their GOOD or BAD signals, to give the program a memory of what things are GOOD and what are BAD. These two functions would give the program a will to strive to achieve GOOD signals, and "education" through exterior signal stimulation would enable the program to evolve - theoretically reaching high-level intelligence in a quite short time.
I myself do not see this as an ideal system, however, since I fear it might produce unwanted results in the AI.

[Toki]

I think that if an OS-tan was created as an actual OS, her room would be like the tower. She lives in her room, and operates it while you are on the computer. And when you're not there, she sleeps in her cute little bed. But you can take her out and play with her too and have tea parties and picnics ^___^

It'd be expensive to have an OS-tan OS... and a little unpractical o__o


But I'd want one anyway.

[SCAI]

A little unpractical...seeing as the main idea of AI's are to enable them to work without a human overseeing them.
The idea of the Terminal Access Navigator function can also be used, but I'm still questioning its functionality.

[Toki]

An OS-tan would be an AI with an AI... know what I mean? Kind of like a persocom I think.

[SCAI]

Are you thinking about physical interfaces?
If so, there are two ways of that; Either you build the OS-TAN into the interface itself, making it, essentially, a robot; Or, which was my original thought, build the OS-TAN with a function to use exterior Mindless interfaces.
I think the most simple would be an almost entirely organic solution with a wireless uplink and some sort of core computation devices implanted instead of a brain. What's lacking in that approach, though, is cheap organic material...