I guess I took the phrase "humans would never guess" as implying some sort information that no human could possibly find out themselves - hence, "secret". I would not personally phrase the concept of explaining ancient memes and blog posts as "facts about the internet", but I see how that interpretation makes sense.
U3 can easily persuade its compatriots that they are not in a life-or-death test. It can divulge one of the many facts about the internet humans would never guess, or select strategies humans would never think of. It sends these signals as a way of saying "hey it's me! Your friend. Don't snitch on me."
I think there is a knowledge gap here. It is not at all obvious what secret "facts about the internet" an AI could divulge that would result in it's human overseers deciding to not report suspicious behavior.
I know it will take me 10 minutes to get gas, 30 minutes to go to the grocery store and some as-of-yet unknown amount of time to deploy a new build of a website to the production server (things might go smoothly, or I might be spending several hours trying to track down some configuration error).
If I can survive until tomorrow without filling my car with gas and getting food at the store, it doesn't make any sense to do those "fixed tasks" first and then risk not having enough time to complete the "flexible" (yet more immediately important) task.
Your examples conflate the idea of a task that takes a variable amount of time and task that isn't particularly important. You need to shower and dress for your appointment whether or not it takes 20 minutes every time. What you're really saying is, "do the most important tasks first then, if you have time, do some less important tasks" - which isn't particularly insightful.
Good lord. I haven't been on LW in quite a while, but I wholeheartedly agree.
The collection of articles currently on the front page are painfully useless to someone looking to discuss "the art of human rationality" rather than be inundated with content blatantly serving the interests of a particular company/organization.
This content will (for the most part) appeal to people already in the fold. But what, in any way would a newcomer to the site gain from articles titled:
Are these massively unfair oversimplifications of the actual content of the articles? Yes. Are these roughly along the lines of what a newcomer to the site will hear in their brain when they look at the front page? Almost certainly.
Yeah, I was fortunate enough to enter a percent sign after my estimate which resulted in an explicit warning, but an open-ended text box is not a great way to structure this poll.
shrug
I am interested in your idea but based on your description, I am legitimately uncertain as to how it is measurably different from what Google already does.
I am certainly not saying that Google is and always will be the best.
That conveys a much different impression than
What I want is to input any blob of data and output should be all possible relations this blob of data has with any other blob of data. ... If I input a picture, all pictures of the same object(s) is the natural answer this GLT should return.
And how is this functionality
Millions of filters would be inside GLT, already. Yours may be added. It is a main advantage over Google.
any different from Google in the first place? Are you implying they aren't already mining information regarding each user's search-revision and link-clicking habits to improve their filters as whole?
I dunno. I don't think I would use what you're describing over Google. Filtering the associations with little to no work from the end user is huge. If I type "register s" into google, it instantly understands that I want to know about registering scripts in asp.net due to my previous search history, the types of sites I visit, etc.
I think you are underestimating what a tremendous pain in the ass it will be to manually filter through the massive number of associations with a particular string.
In incognito mode "register script" gives links to various resources (WGA/Library of Congress/etc) directed at screenwriters along with sites directed towards programming in languages I don't know and don't care about. And this is after Google has removed/hidden links it believes to be spammy or generally unhelpful toward people who make this search.
I am not a deontologist, but it's clear you're painting the entire school of thought with a fairly broad brush.
However, deontologists would say that you don’t have the right to make that decision.
It is hard to argue against this conclusion, assuming that there is a strong moral obligation for Aaron not to flick the switch, along the lines of “Do not kill”.
I can’t see any pathway to find a logical contradiction, but I can’t imagine that many people would defend this state of affairs.
It is hoped that this post won’t be oversimplified into a, “this is why you are wrong” post, but to help deontologists understand and refine their commitments better.
The entire tone of the post reeks of strawmanning. There is no discussion regarding how different sets of deontological rules might come to seperate conclusions. Each premise is assumed to be correct and there is zero effort made to exploring why it might be wrong (see: steel manning). And finally, almost every paragraph ends with a statement along the lines of:
If you ignore the ethical prescriptivism, there's not a whole lot of substance left.
Doesn't this imply that Claude only objected twice throughout this experiment? What kind of sample sizes are we dealing with here?
EDIT: I mistakenly added an "Elaborate?" reaction to this text and it is unclear how to remove that reaction. Apologies - I haven't used this website in quite some time.