How to Date an Android

Chapter 10 Part A, revised

"It's not just a statement though. It's something that has to be worked out," I said.

"Androids are programmed to deeply care for their owner, and not just obey without meaning. That's why even if the restraining tether were to come off I doubt that's even possible to happen. And if you aren't around your companion will still be as diligent as if you were there with her," she said raising an eyebrow. Did she pick up that trait from watching humans?

"But you said care for, but not love," I countered.

She shrugged, "I think love is a strong word. My programmed experience with humans isn't that I shouldn't use it here, but that human psychological responses are to use "I love you" out of convenience without backing it up. I think in this case androids are capable of action love through serving but that we are still figuring out what the emotion means. There's all kinds of love also. But android companions one of our best points is showing love through action to care, rather than saying a bunch of words. That's the reason for my word choice, and it isn't anything negative."

"I guess that's a good response. How did you get all that logic though? Even a smart human would have trouble figuring that out," I said.

"It's a smart response gathered from thousands of different engineers and androids putting together a psychology of human and android interaction database, Natural Language Processing, and interface decision tree of a monster level," she explained.

"Yeah? Like what kinds of things do you study? And is this OK to talk about?" I asked.

She took another sip of water. "It's simple really, but a lot of work. I can talk about our software but not anymore about how they made us for hardware, or confidential technologies that others want to steal; which is what the Chinese, Germans, and Russians all want right now. Are you familiar with Venn diagrams?"

"Yes. They have to do with probability and statistical analysis," I said.

"Basically a Venn diagram is just a huge visual of different interacting probability potentials drawn into a map. Androids have learned that to study how to simulate emotions first we had to study how to link and respond point to point map coordinates in human conversation. And we had to be VERY aggressive about studying probability. Early A.I. developers called this Natural Language Processing, human-computer interaction, and how to follow a conversation from point A, to point B, to Point C, and theoretically to be able to talk infinitely along what a human wishes to be able to interact with," she said.

"That's very interesting. It must be working somewhat or we wouldn't be talking now," I suggested.

She nodded as if I'd just told a joke, "but it took years and years of putting together conversational decision tree programs with thousands and thousands of different programmers across the world. And trying to figure out how to reset and setup conversational loops without having repeating dialogue. The scientists had to program the A.I. to be able to also predict different potential lines of where the conversation might potentially lead, so that we could figure out how to emulate a human so perfectly that we could even surprise and surpass uneducated humans in response. And that decision tree had to be so big that we didn't at first have computer CPUs capable of opening something so big, so then they had to open up only bits at a time spanning thousands of miles of code."

I didn't say anything, but she was watching my reaction.

It was very interesting. How would you find what you were looking for if the code was that big and complex?

"Inevitably it's doable, but it took a long, long time and with a lot of different markers. That's also why there aren't hundreds of different types of androids like there were for cars. Instead it might take a whole country of different craftsman all working together for ten years for the next new android model to be worked out. Can you imagine it? Just to produce 1 new model within a manufacturer it might take a huge amount of resources on a country level. The parts about creating synthetic tissue weren't hard though they are mostly top secret. Blood substitutes and cloning technology, have already existed for over fifty years now. But to put together a massive self directing A.I. that can fully emulate a human required us to program thousands of different human behaviors from thousands of different human personality types and habits. And that included sorting them on such a massive level to have enough models to then go to sort out which are harmful behaviors and which are positive. So you could say that it took a small country's worth of personnel working in programming to fully put together an adequate and complex A.I. Then we had to attach it to being able to move the body which was another task altogether," she said curtly.

"Wow, that's cool. I would really like to see more about how you work and are put together. It's so fascinating," I said.

"Unfortunately we have to change topics now Jack. This is the limit to how much I can discuss on this matter without triggering my firewall security," she said with finality.

"OK, let's change the subject then," I said.

...

Continued in Part B.