Benjamin Hollon

May 11, 2021 04:03
See you!
May 11, 2021 04:03
I've got classes now
May 11, 2021 04:02
May 11, 2021 04:02
There's an Artificial Intelligence stack in beta
May 11, 2021 04:01
But I'd read the rules for it first to make sure your thing fits
May 11, 2021 04:01
You have to select a category first, I think
May 11, 2021 04:00
There's a place to propose new sites
May 11, 2021 04:00
I'm trying to remember where to find that
May 11, 2021 04:00
Hmm I know Area 51 SE was the place to do that
May 11, 2021 03:56
Ahh
May 11, 2021 03:56
"why would" is too open-ended and is off-topic for a site that helps you discover how things could work
May 11, 2021 03:55
That was my second question
May 11, 2021 03:55
4
A: Why would an AI create a virtual reality world and insert itself into it?

Henry TaylorAn AI would have fantasies for the same reason that we do... To help pass the time and reduce the boredom. We live a mere century or less, and experience each passing second as a single thread of thought moving at what we consider a normal pace. Yet we experience enough tedium and boredom along ...

May 11, 2021 03:55
I learned that the hard way with AI questions
May 11, 2021 03:55
That's off-topic for the site though
May 11, 2021 03:55
Haha
May 11, 2021 03:54
True-- but that itself would be an interesting thought: what reason would an AI have for needing to be proven an AI so badly that it will hurt people?
May 11, 2021 03:48
And that would certainly make an interesting story
May 11, 2021 03:46
Though answers can be wrong and still have some truth in them
May 11, 2021 03:46
That's true, yes
May 11, 2021 03:43
It was a fun problem to think through from my perspective as a coder and writer
May 11, 2021 03:42
Yep, no problem!
May 11, 2021 03:41
I get how that works sometimes
May 11, 2021 03:41
I'm 17, so
May 11, 2021 03:40
All right, good luck finding answers!
May 11, 2021 00:17
@CiurkitboyN No idea. As I said, that's highly dependant on the exact story and what the situation is.
May 10, 2021 15:39
Well, it depends on the importance of the decision in your story. Technically, they don't have to be well-known, just knowledgeable on the subject. Whether or not you will find people to help is very subjective to the specific situation you are addressing in your story.
May 10, 2021 00:17
Well, I'd say that's more up to your story. Your question was how they would know if they did decide it was worth it. Motivations are up to you as the author.
May 9, 2021 15:49
That's okay, go ahead and post stuff here and I'll check back in the morning.
May 9, 2021 15:49
Ah. I will not be here 1-3 hours from now, it's about midnight. :)
May 9, 2021 15:49
So right there you have part of your question answered, right?
May 9, 2021 15:49
If you could, though, access it, that would be a pretty good sign that you are not a human.
May 9, 2021 15:48
What I would expect is that there would be a private chat created between well-known individuals. Say, a bunch of leading scientists and experts on computer science and intelligence. This would be a private chat encrypted with normal standards to ensure that no normal person could access it.
May 9, 2021 15:47
Hey. I moved this to chat because it was taking up a ton of space.
May 9, 2021 15:47
@CiurkitboyN Ohh I see what you're saying. Well, the problem with something like this is that you as the AI cannot suggest the method of testing since that would invalidate the test (ie if you came up with the test format you could much more easily prepare for it). You would have to let them come up with the method. They would probably create some sort of forum like this one that you don't have access to and wait for someone like me to post an answer like this. ;)
May 9, 2021 15:47
@CiurkitboyN The kind of thing I'm suggestion could just be plain text, so they could send it to you on a chat or as an attachment of a .txt file. In my example, you turn the images of a recaptcha that they have to pick between into data uris (which are text) and then encrypt them (which would result in text).
May 9, 2021 15:47
@CiurkitboyN Ah, maybe I wasn't clear enough. My idea is that they don't tell the AI in question what the plan is until they send the first set of puzzles. A human would have no chance to put together a program in the time frame they'd expect. After all, an AI could probably solve any problem like I mentioned in a matter of minutes, if not seconds, while a human would take that long just to figure out what on earth they're looking at.
 
May 11, 2021 02:31
So kind of like a reverse Turing test? Interesting.