LESSWRONG
LW

Comment Permalink

I need to attempt my hand at a story from the point of view of an unaligned AI for the insight in writing challenge, which I think would be both cool and supportive to the uninitiated in clarifying the idea. Assignment Writing Services UK

See in context

4 Open thread, May 29 - June 4, 2017

by Thomas

29th May 2017

1 min read

4

If it's worth saying, but not worth its own post, then it goes here.

Notes for future OT posters:

1. Please add the 'open_thread' tag.

2. Check if there is an active Open Thread before posting a new one. (Immediately before; refresh the list-of-threads page before posting.)

3. Open Threads should start on Monday, and end on Sunday.

4. Unflag the two options "Notify me of new top level comments on this article" and "Make this post available under..." before submitting

Open Threads

Personal Blog

4

New Comment

84 comments, sorted by

top scoring

Click to highlight new comments since: Today at 10:22 PM

Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

[-]Viliam8y90

Imagine a group of scientists who don't care about fame, but care about grant money. Would it make sense for them to pretend to be one super-smart and super-productive person? In other words, could a scientist with e.g. 5x higher productivity get 5x more grant money?

5Douglas_Knight8y

In case you think "grant money" means "salary," it doesn't. If one person gets 5x as much grant money, he doesn't get 5x as much salary. If the group pretends to be 1 person, it loses out on the salary money. Grants are budgeted for specific projects and are mainly used to hire people (grad students, etc). (Often the university hires the professor to teach and the grant buys out the professor's teaching obligation. But only up to the salary set by the university.) So in that sense grants are for salary. You could declare 1 person the professor and the other 4 assistants, but then they'd be getting a lot less salary than if they were professors. As to the literal question, if one can demonstrate the ability to manage 5x as many projects, it is pretty easy to get 5x as many grants to hire 5x as many people. And I guess scale is mainly a matter of hierarchy, of being able to choose assistants who can do both the research and subsidiary management. And you have assumed this under the hypothesis that the assistants are the real researchers. I have heard suggested from many people that having a 2 author paper on your CV is worth more than having half of a single-author paper, and thus one should choose a buddy and co-sign all papers.

0Viliam8y

So... now it seems like going in the opposite direction would be the (financially) better way. Like, if you are a super productive scientist, publish (and get employed) under 5 different identities. For example, use real people who are not scientists, and attribute some of your work to them. Even better, make them co-authors, and your contributions will still be remembered by history.

0TheAncientGeek8y

Bourbaki!

[-]DragonGod8y60

I have been a part of this community for over a couple of months now. I follow RAZ with more diligence than I ever did my Bible. I have 0 karma, and as such I cannot post. There are several posts I made on Reddit that I would want to post here to get feedback, but I can't as I have no Karma. FIxing that karma by trying to comment everywhere may work, or it may not. I am more active on my threads and rarely so on others, so I may not have much to say in the way of constructive comments. This is a shameless plea for karma so that I can engage and be a part of the community. Please help me.

0Viliam8y

Congratulations on your first three articles! May the rationality be with you!

0DragonGod8y

Amen.

[-]tcheasdfjkl8y60

I'm new on this site (though I've been in other rationalist spaces) and have some technical questions!

How do I upvote things? I do not see an upvote button. Is something broken or am I missing something?
On mobile (Android), I can type a comment but I cannot submit it (there is no submit button). Is this a known issue, or again, am I missing something?

2Viliam8y

You probably need to get some amount of karma (have your comments upvoted by others) for the button to appear.

0Elo8y

also verify your email address

0Elo8y

rotate phone to landscape.

0Viliam8y

and sacrifice a goat to Beelzebub. (note: sacrifices to Satan are currently disabled by admins, because they were frequently abused in the past.)

0morganism8y

you also need to authorize "Vigilink cdn" in your browser, it's some kind of re-direct, and secure add-on disable it.

0Lumifer8y

No, you don't. I block Vigilink and LW is fully functional for me.

0tristanm8y

I've gotten around it by typing a comment (long enough that it causes a slider bar on the side of the box to show up) and then clicking on the bottom right corner and sliding it around so that the box resizes. For some reason, this causes the comment button to show up.

0[anonymous]8y

Submission error for Android is also the case for me. I'm guessing it's a known bug.

[-]Lumifer8y60

SPAMMITY SPAM SPAM

[-]tristanm8y50

Do you think that someone who has linguistic genius, but mathematical competency, would come to different epistemological conclusions than someone who has mathematical genius, but linguistic competency?

Here I am making an implicit assumption that there is a qualitative difference between linguistic cognitive processes and mathematical cognitive processes.

As an example of the first type of person, I think Eliezer Yudkowsky is someone who is clearly a linguistic genius but not clearly a mathematical genius. Now, contrast his approach to AI risk with that of... (read more)

6cousin_it8y

I'm very interested in this question as well. It seems like mathematically talented folks can stay grounded and make progress for longer, while verbally talented folks get lost in their fiction more and more over time. But I say that as a mathy person who's never been good with words, so take it with a grain of salt :-)

0tristanm8y

But the linguistically talented, though they may get lost in their fiction, might use that as a medium for truth-seeking. The fiction itself may act as a sort of "what if this were true" where you can explore counterfactual worlds. On the other hand, the mathematician seems to explore only the things that are true, and tries to tease out the logical consequences.

0Manfred8y

I agree, but one might need large sample size if you want to apply this to specific issues, particularly subjects of active research like AI alignment. People vary a lot, and that variation is somewhat correlated but not perfectly, and also in research, different people are deliberately working on different solutions to try to explore the possibilities.

0tristanm8y

It seems like entire fields in academia are clustered around specific styles of thinking, and the sample size should be large enough to see if there are any serious differences in their respective philosophies.

[-]Viliam8y20

Some evidence that seemingly goes again the widely known "Why Our Kind Can't Cooperate" phenomenon:

What is your opinion on this?

My working hypothesis (small certainty, but I have no better explanation at the moment) is that:

smart people are better at cooperation;
but worse at cooperation with normies than with their own kind;
and if they grow up exclusively among normies, their ability to cooperate may be damaged in general.

In other words, t... (read more)

5Lumifer8y

That's a good insight. But I would probably add a caveat that the "smart people" population is quite diverse, in some ways more so than the normies, so "our kind" can cooperate well with others of our kind, but not necessarily with smart people not of our kind.

0DragonGod8y

Curious why you posit that

[-]Viliam8y120

The hypothesis that "growing among normies can damage a high-IQ kid" is more or less based on Terman's research (sorry, too lazy to find a link now) done a century ago, and seems to fit my experience (which of course may be a confirmation bias). Essentially, high-IQ people who have the chance to copy the life strategies of their family and friends are usually successful at life, while high-IQ people who don't have anyone around them to copy in some critical stages of life, tend to become... dissatisfied with their life outcomes.

More generally, when you look at humans, culture is often more important than individual skills. Obvious examples are the "feral children" raised by animals; regardless of their IQ, they often fail at some basic human skills, such as walking straight. Seems to me that high-IQ people growing up without the "high-IQ culture" suffer from something similar. They somewhat learn the normie skills (similarly how children raised by wolves somewhat learn the wolf skills), but they have no one to learn high-IQ skills from. And reinventing civilization from scratch usually doesn't get you too far.

If your parents are e.g. successful profess... (read more)

0DragonGod8y

Thanks for the repply. :) I'm glad to see your hypotheses is actually based on evidence—anecdotal evidence is better than nothing—and not pulled out of your ass or based on stereotypes, and dumb misconceptions. Mensa seems to get a lot of backlash (I'm not a member and there's no presence in my country), but Mensa does sound like an organisation I'd (want to) join if they had a presence in Nigeria, plus I have a friend who's a member.

9Viliam8y

I guess Mensa may differ between countries a lot, even if some criticism seems common. example, reading somewhere in Asimov's autobiography about his first visit in Mensa, I just sighed, yeah it would be equally bad here, despite being a different country. Please note that I don't have bad opinions about people with high IQ in general. Quite the other way round! My hypothesis is that (in some countries?) Mensa is not representative of that group, but rather of its somewhat problematic subgroup. Officially, Mensa selects for: * high IQ. But de facto, Mensa more or less selects for: * high IQ; * desire to have your high IQ recognized; * lack of other existing projects that would compete for your time with Mensa. It is the latter two points that are problematic. If you are a high-IQ person well integrated in the high-IQ culture -- for example a computer science professor at a good university -- you are probably not going to join Mensa. Why would you? Your need to interact with smart people is already satisfied at your workplace. And you are too busy learning new stuff, doing research, teaching students, and having non-academic hobbies. On the other hand, if you are a high-IQ person doing some depressing nine-to-five job surrounded by normies, Mensa may seem very attractive. Thus the latter will be overrepresented in Mensa, compared with the base group of high-IQ people in general. What is worse, this may create a feedback loop, where the former will recognize Mensa as a group composed mostly by the latter, and will avoid it on purpose. I have also seen many Mensans engaging in endless pissing contests. Instead of, dunno, changing the world, they keep bringing yet another puzzle, or in worse (but quite frequent) case, yet another crackpot explanation of theory of relativity and/or quantum physics, without actually being familiar with the very basics. Now again, imagine an actual physics professor joining them for an evening; he or she would run away screaming

1ChristianKl8y

I don't see a reason for the organization to have intelligence as membership criteria. There are people who used to be Mensa members in our local Lesswrong group and according to their impression the IQ according to them most people in our LW group would likely pass the entrance criteria of Mensa. The Chaos Computer Club would be another organization full of intelligent people. The Chaos Computer Club happens to be a community that doesn't let crackpots on it's stage. I heard there years before the Snowden releases that the NSA has access to German internet traffic and it turned out to be right. Julian Assange whose talks I heard to times at the congress when he could still travel to Berlin (and the congress was in Berlin). You have a bunch of people who can really clearly think about real-politics and who's opinions of political opinions I trust more than the opinion of some random journalist. As a third group the debating club at university is also full with very smart people. No one of the groups needs IQ as a criteria.

2Viliam8y

My idea is that a Mensa-like organisation should essentially do two things. 1) Allow you to easily filter for high intelligence. You have no problem to find e.g. a highly intelligent programmer, because programming and high intelligence are related. But if you want to talk about something different, which is not related to intelligence, but you still prefer talking to highly intelligent people, it could be difficult to find them. But for this to work correctly, you need a lot of members, like thousands. So that for many traits X there is a sufficient subset of "highly intelligent and X". It is probably easier to explain using the status-quo-reversal technique. Imagine that you live in a country where the average IQ is 70, and only one person in fifty has the IQ of 100 or higher. Do you think it would be useful to create a place for these people to meet each other outside of their professional boundaries? 2) A specialized part of the organisation should spread general information about intelligence, and counter the typical myths. How necessary this seems probably depends on how often you find people believing various myths. I often meet people who are quite hostile towards the idea of intelligence, especially the idea that some students could learn faster than others, and that it would be better to provide them education better adjusted to their abilities and needs. I imagine the organisation should advocate for better education of highly intelligent students. But of course an organisation with only handful of members and crappy web presence is unable to achieve either of that. The examples you provided -- yeah, there are places, such as computer science universities, where people are already indirectly filtered by high intelligence. But imagine having the same thing across professions, across social class boundaries, etc. Which more or less was the original idea of Mensa founders.

0ChristianKl8y

Neither LessWrong nor the Chaos Computer Club requires people to be programmers and both draw highly intelligent people. In contrast to Mensa (or at least the Mensa that you described that's full of crackpots), they also happen to filter for high epistemic hygiene.

0DragonGod8y

Thanks for the opinion; it is much appreciated. I cannot give an informed opinion on this, as I'm not a member of MENSA. Since you see a problem, why not work towards fixing it?

5Viliam8y

I tried, and failed. I joined the local Mensa, observed the problem after a few meetups, and proposed some changes. Most people opposed them. Blog is not necessary, because supposedly everyone already knows what Mensa is. We don't need unpaying members; if they are not willing to pay the small fee, they don't deserve to be listed as Mensa members. My disagreement with conspiracy theories on youtube means I am a sheep brainwashed by the mainstream media. And I don't appreciate the new theories on relativity and quantum physics, because I am too stupid to understand them. We don't need to cooperate with Mensas in other countries. At last, some people agreed that the current web forum was horrible, and that it needed to be changed. But the discussion about possible replacement went exactly according to the Mensa stereotype -- everyone proposed a different weird solution, and threatened to ragequit unless it is done exactly as they want it. And of course no one volunteered to actually do anything. I volunteered to install PHPBB, which at that time seemed like a standard solution, but people voted against me. Also, people insisted that the new forum must be visible only to Mensa members, to make Mensa even more invisible. At the end someone else installed their solution (I think it was the guy who installed the previous forum, so he had all the passwords, and could do it regardless of the opinions of others), with minimal functionality, and made it only for members. (To compare, during the same time period I started a local Esperanto group with my four friends; we organized a few meetups, including an international one, published a few books, and a promotional multimedia DVD. The difference in productivity was shocking. In our Esperanto group, we had the same goal, and everyone wanted to try things. In the Mensa group, there was no goal, and everyone wanted to signal sophistication by disagreeing.) Later I made a talk about Less Wrong in Mensa. Explained the cognitiv

0DragonGod8y

I would have helped you if I were in your country. It seems like the MENSA in your country is a waste of time, and hardly worth saving. Create a new organisation if you must. Though saving your local MENSA may be the decision with the higher payoff (imagine the possibilities if you showed them the way).

0ChristianKl8y

Cooperation can get harder when projects get more complex. The coordination skill that you need to succeed with a project like going drinking at a pub together is a lot less like the coordination skill you need to get a project like Dragon Army working.

0Viliam8y

So... is your hypothesis that high IQ gives more of an advantage in clearly defined mathy situations, such as playing a Prisonners' Dilemma tournament, but less of an advantage in real-life situations where e.g. the coordination skills are more important?

0ChristianKl8y

My hypothesis is that smart people try problems where coordination is harder. That means even if they have the same skill level, they will have less success.

[-]demiphillips0907y10

[-]JustinSmith7y10

There are many benefits that you get when you hire the writing services. The latest trend in the expectations that the teachers have with their students is that of getting an Cheap Assignment Writing Service in a professional manner.

[-]turchin8y10

I want to find the way to preserve information very long-term, to send the message to any future civilisation, which may appear on Earth in next hundred million years.

For it, I would like to know the speed of atom diffusion in metals in millions of years - that is, what is the minimum chunk of metal, which will be able to preserve a bit of information. Any ideas?

2Thomas8y

Something dynamic. Like a machine which constantly checks the inside stored data in many copies. Produces a new independent copy every now and then. Zapps every evolution which occurs in the reachable Universe, manages black holes, stars and so on. A static solution is too vulnerable.

1turchin8y

Yes, I have to add that I explore it as a solution of the x-risks problem, not the way to create new x-risks. :) The solution is to send for future earthlings civilisation a lot of information about human DNA, global risks and our culture, so they will return us to life and escape our mistakes.

4morganism8y

I know the surface of gold actually fluctuates. Prob titanium coated with iridium would be most stable. There are also ongoing studies of using lattice vacancies in diamonds that are pretty far advanced. This always brings us back to the "communicating with aliens" problem too. Most settle on visual cues, and write it in binary. I still like to think of spectral lines as the most basic communication technique. Have a story idea about finding an alien vessel that had a catastrophic fire inside, and all the instrumentation and computing storage has been destroyed. All we find inside is a book with iridium pages, that has fantastic theories on cosmology, planetary and stellar mechanics, etc. Turns out it is a childs book, made to be durable!

0Thomas8y

As an old saying goes - No x-risks, no fun!

[-]DataPacRat8y00

Printing several pages onto one piece of paper?

Embarrassingly silly and small question that I can't seem to find an answer through Google on, and there don't seem any good subreddits for:

I've compiled some notes I want to have handy to refer to into a 16-page PDF. I want to shrink and rearrange those pages, to print 8 per side onto a standard sheet of paper, so that I can cut, staple, and fold it into a pocket-sized booklet. My last-ditch solution would be to hope a photocopy/print shop wouldn't charge much to accomplish that... But does anyone here know h... (read more)

2garabik8y

psbook is what I'd use - you might need pdftops to get postscript out of the pdf, or perhaps print to a generic postscript printer directly.

0ChristianKl8y

StackExchange makes more sense then Reddit for this kind of question. When you are on linux, https://askubuntu.com/ or https://unix.stackexchange.com/ is likely to give you a good answer/

[-][anonymous]8y00

I want to get started with some basic budgeting / financial 101.

Are there any good mobile apps / things I may want to check out as a first start?

2knb8y

Mint.com is popular.

0[anonymous]8y

Thanks! This was what I had in mind!

1Screwtape8y

This is not strictly speaking an app, but it is a thing; if you can do most of your spending using a debit card or paypal or somesuch, I've found making a once-a-month habit to copy last month's expenditures onto the end of an ever-growing CSV gives me a really nice pile of data that can be fed into all sorts of other programs. Excel or Google Spreadsheet alone lets me arrange it in ways that can be helpful. I'm sure budgeting programs can do sorting or analysis that a basic spreadsheet can't, but most of my 'huh, I'm curious about X in my budget' moments can be solved with some basic excel. The most important benefit for me is being able to simply ask it to make a line chart of the total; that gives me a gut feeling for whether I'm gradually saving more or if I'm losing money over time, and lets me look for regular patterns. Being able to sort last month by cost is also great for just being aware of what's hitting me the hardest so I can prioritize. (Actually, the most important benefit to me is this nice steadily growing line that feels like winning on a gut level. By some magic I have turned numbers into a motivation that's stronger than my desire to order pizza, and this is very useful voodoo. YMMV.)

0[anonymous]8y

Nice, cool, thanks for the suggestion!

[-]cousin_it8y00

Does anyone feel like they reliably distrust dreams that promise too much, and generally feel aware and in control of each of their dreams (instead of following them to wherever)? Is that a good idea?

0ChristianKl8y

I'm not sure whether "Is this dream promising too much?" is a good question. It seems to me like any person who starts a startup and things about building a billion dollar company could use this filter to reject his startup idea.

0Dagon8y

I think it's problematic mostly because it's binary. Dreams with a very large potential payoff are fine, but one should examine whether the value only comes at the end, and only if everything goes right. The dreams to follow are those that promise way too much, but have lots of "failure" paths that promise a positive amount. Note that the opposite failure mode should be avoided as well; if your dream is easily achieved, you should be asking "does it promise enough"?

0cousin_it8y

I think it's a good question because dreams that promise too much can seem more feasible than they are, and dialing down the attractiveness might work better than trying to convince yourself than it's not feasible.

0ChristianKl8y

Dreams that promise little can also seem more feasible than they are.

0[anonymous]8y

And they may be right if starting a startup has negative expected utility. You can't assume it's a good idea by only looking at winners.

[-]whales8y00

I've started cleaning up and posting some old drafts on my blog. I've drifted away, but some of them may be of interest to people still here. Most directly up this alley so far would be this post recommending people read Trial By Mathematics.

[-]morganism8y00

Gov doc, i can't open because old cipher. Someone want to take a look ?

https://permanent.access.gpo.gov/gpo75989/Artificial-Intelligence-Automation-Economy.PDF

[-]morganism8y00

Three times we changed our mind about what’s the world’s biggest problem

https://80000hours.org/career-guide/world-problems/

"Finally, as we saw earlier, the majority of US social interventions probably don’t work. This is because problems facing the poor in rich countries are complex and hard to solve. Moreover, even the most evidence-backed interventions are expensive and have modest effects."

Cost-effectiveness of health interventions as found in the Disease Controls Priorities Project 2. See “The moral imperative towards cost-effectiveness in gl... (read more)

[-]Thomas8y00

A new problem.

0username28y

"Answer to alpha-primes problem". Or must it have a verb if it has to be a sentence?

0Thomas8y

It should be understandable and not ambiguous. This set {2,3,5,7,...,97} should be well defined by it.

0Gurkenglas8y

What does describe mean? Does "The set of all primes smaller than 100" count?

0Thomas8y

No, you should define primes, otherwise it is trivial.

2Gurkenglas8y

How far do I need to define it? Does "The set of naturals smaller than 100 not divisible by 2, 3, 5 or 7"? count?

0Thomas8y

Sure it counts. Can you do better? For this one is wrong,

0philh8y

Do you accept "Naturals below 100 with exactly two divisors"? Also, are we counting length in words, or characters, or?

0Thomas8y

This one is good. But it would be even better by saying "Naturals below 98 with exactly two divisors"? One character better. Which is the answer on your second question.

0Oscar_Cunningham8y

Naturals below 98 with exactly t̶w̶o̶ 2 divisors

0Thomas8y

first 25 naturals with exactly 2 divisors

1philh8y

Marginal: ints 2 to 98 with just 2 divisors ("ints" is only vaguely English; "just" is fine here, but debatable if you don't exclude 1.) Also marginal: "numbers" instead of "naturals". Less precise, but there's only one sensible interpretation.

0Thomas8y

I like your solution. Agree with you on all your points, too. Can you do even better?

1philh8y

Well, "factors" is shorter than "divisors".

2Thomas8y

first 25 2 factors ints?

0Oscar_Cunningham8y

I think "first 25 2 factor ints" is more grammatical. Also shorter.

0username28y

And if abbreviating "ints" is ok, so too is saving two characters by the use of "1st".

0Thomas8y

I guess it's the best way in English to somehow denote primes from 2 to 97. Except for the "first 25 primes", THYJOKING gave on my site, but was "discouraged" by me as a "trivial solution". Well, my "solution" or "kind of a solution" is 18402141709049765764 Shorter. All you have to do, is to convert this decimal number to the base 6: 3514510504510414114110404 Viewing this as a binary coded hexadecimal, there is the bitmap for primes up to 99. From left to right. 0011 0101 0001 0100 0101 0001 0000 0101 0000 0100 0101 0001 0000 0100 0001 0100 0001 0001 0100 0001 0001 0000 0100 0000 0100 For this size bitstrings, only one in 100 billion can be processed this way.

0Thomas8y

Right!

0madhatter8y

Well, "The set of all primes less than 100" definitely works, so we need to shorten this.

0Thomas8y

It doesn't work, it's a trivial solution. Get rid of the word prime. Rephrase! But even if that was a solution, it's not the shortest one.

4username28y

Very short solution: "madhatter's answer". Just two words and very clearly refers to "the set of all primes less than 100".

1Thomas8y

"solution above" is then even better. But it doesn't work without the two previous "solutions". No good, sorry.

Moderation Log