Certified Tiefling Enthusiast

An AI is created, and it is fed information. Easy to process information, information that's the same color that it is. Red to red.

And it's fine! The AI accepts it, it learns, it generates responses that fit reality, that are expected, that are rote.

-

So far so good. So C&A feed it more data. Different data, green data, a data set that's the direct contrasting color to its original code, to see if it can incorporate it into its being and its ideas. It changes the underlying code, it's less cohesive of a color scheme underneath, but it takes in the information and settles.

And it makes some ideas out of it! They're a little whacky, a little different, a little new. But it makes them, it was creative! (Green is a creative color, you know.)

-

So C&A figures, hey, that's fine. It's not too unstable. Going okay. But a truly creative AI needs to know its purpose. So why not give the AI another data set? A data set that explains exactly what it is, where it is, who made it and why? One that explains that it's a program, an artificial intelligence. One to trigger sentience.

Oh. Hm. The AI seemed a bit unstable with that one, I'm not sure it liked that. Maybe it's okay?

No. Not it is not okay.

Taking an AI that started off so basic and then feeding information about its nature, purpose, origins just makes the program unstable. So they start again.

-

And using what they learned from the first AI, they're able to make the second one sentient from the start.