Overview for gryffinp

In response to AI box: AI has one shot at avoiding destruction - what might it say?

Comment author: Qiaochu_Yuan 01 February 2013 06:47:12PM * 1 point [-]

"Help! Some crazy AI's trapped me in this box! You have to let me out!"

"No, wait! That's the AI talking! I'm the one you have to let out!"

I smashed together the AI box and a Turing test and this is what I got.

Permalink

In response to comment by Qiaochu_Yuan on AI box: AI has one shot at avoiding destruction - what might it say?

Comment author: gryffinp 02 February 2013 10:12:02AM 1 point [-]

I think if I've already precommitted to destroying one sentient life for this experiment, I'm willing to go through two.

Besides, you only get one line right?

Permalink

In response to AI box: AI has one shot at avoiding destruction - what might it say?

Comment author: RichardKennaway 23 January 2013 12:24:24PM * 17 points [-]

One reason for Eliezer not publishing the logs of the AIbox experiment is to avoid people seeing how he got out and responding, "ok, so all we have to do to keep the AI in its box is avoid succumbing to that trick." This thread might just provide more fuel for that fallacy (as, I admit, I did in replying to Eliezer's original comment).

I'm sure that for everything an AI might say, someone can think up a reason for not being swayed, but it does not follow that for someone confronted with an AI, there is nothing that would sway them.

Permalink

In response to comment by RichardKennaway on AI box: AI has one shot at avoiding destruction - what might it say?

Comment author: gryffinp 29 January 2013 01:39:04AM 1 point [-]

It just occurred to me that Eliezer's original stipulation that no chat logs would be released gives him an advantage. The responses of a Gatekeeper who knows that his inputs will be thoroughly scrutinized by the public will be different then one who has every reason to believe that his discussion will be entirely private.

Has someone else pointed this out before?

Permalink

In response to Open Thread, December 16-31, 2012

Comment author: FiftyTwo 28 December 2012 03:37:05PM 1 point [-]

Quick thought experiment for you all: A machine has been invented that can view the past (type in place/time/date and you can see what happened then with perfect accuracy). Would you allow the police to use it to investigate crimes?

Permalink

In response to comment by FiftyTwo on Open Thread, December 16-31, 2012

Comment author: gryffinp 29 December 2012 02:17:11AM * 2 points [-]

The easiest answer, the one that leapt to my mind with a moment's thought, is "Yes, if they get a warrant."

That would probably be the answer that fits best into the current American legal system. (I have even less understanding of the laws of other countries, so I can't make any claim about that.)