un: jaycetalis | text

[hexcope]

[So getting used to this new kind of technology was certainly an… experience. He still isn’t really comfortable with the idea of a video recording of himself, and even an instantaneous voice transmission is beyond the level of tech he’s accustomed to so text is the way to go. Unfortunately, he’s using this like some sort of magical LinkedIn, so it’s a very formal letter style he’s got going on.]

Hello,

My name is Jayce Talis. I’ve noticed there’s a few buildings here that seem to be laboratories. If any have availability to lend a room or two, I have research and device development I would like to work on. Ideally if there’s any access to a forge of some sort, that would be preferred, but not required. I have a personal reference if necessary to demonstrate good character.

[He notably does not say anywhere that there will probably be explosions.]

Thank you for your time.

Comments

[hexcope]

True.

No. It's just a piece of equipment, isn't it? The earpieces are tools, like any hammer or wrench.

[ornithologist]

It is. But things get considerably more muddled when a tool becomes capable of holding a conversation.

[hexcope]

It begins to give the sensation of talking to another person. Is that what you mean?

[ornithologist]

It's a classic test of machine intelligence. If you can't tell if you're talking to a person or a machine, does it really matter if it's a machine?

There's issues with that test, of course, but it's a useful thought exercise.

[hexcope]

I understand the point of the test seems to be catered more toward developers of artificial intelligence as a mark of accomplishment, but isn't it clear that the answer is still yes?

[ornithologist]

A strong position for someone who just learned the subject existed an hour ago.

[hexcope]

Is it? Why would someone argue otherwise?

[ornithologist]

The most obvious objection is that it depends on what task we're talking about. Do you care if a machine rather than a person tells you what the weather forecast is?

[hexcope]

No.

[That's no different than just looking at a thermometer, after all.]

[ornithologist]

Indeed. There are some tasks a machine is better suited for, and some they're less suited for. That's a gross simplification but that's what it amounts to.

[hexcope]

Of course. Machines are better than people at tasks like measuring factors and calculating results. Faster results with more consistently correct answers. But I'm also not asking my weather vane how it's day went, or expecting it to do the same to me.

[ornithologist]

It's not particularly pertinent to the weather, no. But there may be other scenarios you'd like computer assistance with that you'd appreciate a more nuanced understanding of human values.

Sometimes we're not looking for a "correct" answer but the "best" one, where we define best through complex ideology and perspective.

[hexcope]

I'm not following. If a computer can't experience "caring," then how would it be able to assist in giving me the "best" answer for a situation that required understanding human values? Wouldn't it need to experience those emotions to understand them, and then translate that into an answer befitting the question?

[ornithologist]

I didn't say it couldn't experience caring, I said it was extremely difficult.

A fully realized artificial intelligence, which presumably Aurora is, will necessarily approach or even achieve sentience.

[hexcope]

But how?

[He just can't find himself agreeing with that. But maybe he just doesn't know enough.]

Is it like evolution?

[ornithologist]

Return to the earlier thought exercise. What is caring? If you can't tell the difference between a machine and a human caring, is that good enough?

[ Harold is not beating the professor allegations any time soon, that's for sure... ]

[hexcope]

[You really aren't, Harold!!

There's a longer silence from Jayce, mulling things over.]

I need to research more on this.

[ornithologist]

[ That's a very sensible answer, which pleases him. ]

There's books available on the subject at the library, though I'm afraid they're some years out of date from my perspective. I'd refer to books labeled with the code 006.

[hexcope]

Any on programming too? For a beginner.

[ornithologist]

That would be section 005. I suggest "A book on C" by Kelley and Pohl.

[hexcope]

Noted. Thank you.

[ornithologist]

I appreciate the interest. If you have any resources to share for me to learn about your scientific approach to magic, I'd appreciate that as well.

[hexcope]

I mean, what's my alternative? Not being interested?

[Sounds like Harold's asking water to not be wet.]

Probably none in the library, unless somehow my own notes wound up in there. My partner and I are the sole developers of this specific type of research in my world. There's other worlds where people use similar techniques though. Gorgug, for example. He's something called an "artificer" in his world, and he's got a few pieces of equipment that run on a similar principle to what I had manufactured, just with a different power source.

[ornithologist]

Is everyone you know interested in your major intellectual pursuits?

[ That sounds utterly impossible. Water is wet and it's also completely disinterested in arid deserts. ]

I'm familiar with Mr. Thistlespring. His work is fascinating, but completely unfamiliar to me. If you feel your work is substantially similar perhaps I should request his high school textbooks from him. [ It would make sense if magic-science was a common factor in universes other than Harold's, the way the laws of physics must be commonplace, but perhaps skip a few iterations. That's reassuring in a certain way. ]

[hexcope]

Yes.

[Is it still ego if it's true?]

But maybe more in the results than the whats and hows.

His textbooks would be similar enough that it'd be a start. Could I also borrow them? Just to compare the similarities.

[Bro, calm down, this is now like twenty books on your to-read list.]