• Date:3rd of July, 828 A.S.
• Recipient:Lane Hakers
• Sender:John Crow
• Source:Battleship Hood, Dublin system
• EncryptionHigh
My greetings.
I am John Crow, Freelancing pilot of "Morning.Hour" fighter. I represent none of the factions that you have business with, though it doesn't mean I don't have an interesting proposition for you.
Had a little project baking in my head. However, lacking skills sufficient to put it at least on a piece of paper, had to ask around about someone who can. Figured out it could be you - since according to my sources, you are no match in terms of software cracking and development, that also doesn't have unnecessary staples of the House Laws.
I'm talking about developing an order-abiding AI here (with self-awareness out of the question), that can pilot a freighter or transport-sized craft and utilize both known Sirius sector charts and marketplaces information to trade goods with optimal revenue. Maybe with some robotic workers to automate workplaces and a simple сhatbot, based on a recurrent neural network, with voice simulation and synchronous translator - to fool annoying House police or a pirate, if they decide to check craft's insides. Everything should be running autonomously (and put to a stop,if needed), installed on a ship, excluding log servers, that could be placed everywhere you like.
On my side I'm ready to fund the whole project during its progress and when it'll be put under my command upon completion, to wire a significant cut of incoming credits to the agreed bank account.
Should we sign a contract to make this venture more official, or in case if the implementation of such project will be deemed impossible - please, let me know.
The type of project you've outlined would cost at my first thoughts, in the scale of trillions of credits. Not millions, or billions, but trillions. It's a thousand billions, billions being a thousand millions. I'm not sure if you've heard of such numbers before. You see, if you had any hope of funding such projects, and I guarantee these are multiple projects, we would have heard of you. Were it so much as possible you might fund this, we'd be paying you considerable attention already.
But I have not heard of you. So, briefly, I'll outline some more reasons that this isn't feasible. An artificial entity incapable of learning, and only following orders, would if left to function autonomously, require the ability to problem solve. Problem solving is a constant requirement in the void of space. This is an extremely complicated endeavor alone which would require enormous computing power and hardware invested into a ship.
Perhaps you haven't been in so wealthy an area of space to see one, but simple workers and robots with vocal synthesizers exist. They manage simple, repetitive tasks, with little to no need for problem-solving. They are however, temperamental and inflexible. In other words, unsuitable for remote, autonomous operation. These robots cannot answer something they haven't been programmed to anticipate. Imagine these being intercepted and interrogated by corporate militaries or righteous liberators of wealth. Although I do imagine some simple pre-programmed responses or basic chatbots might easily confound Liberty's finest... well, something for me to think about.
In any case, I encourage you to first save for one of these existing robots. Try to get it to fly a cheap ship you don't care for. At best, it will result in something entertaining for the neural net. At worst, it will somehow result in your demise. That may also be entertaining for the neural net. And before you ask, remote control is unfeasible at distance due to signal speed, degradation and interference.