Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Archimedes's Chronophone

20 Post author: Eliezer_Yudkowsky 23 March 2007 05:43PM

Think of how many generations of humanity would have benefited if certain ideas had been invented sooner, rather than later - if the Greeks had invented science - if the Romans had possessed printing presses - if Western civilization had turned against slavery in the thirteenth century.

Archimedes of Syracuse was the greatest mathematician and engineer of the ancient world.  Imagine that Archimedes invented a temporal telephone ("chronophone" for short) which lets him talk to you, here in the 21st century. You can make suggestions! For purposes of the thought experiment, ignore the morality of altering history - just assume that it is proper to optimize post-Archimedean history as though it were simply the ordinary future. If so, it would seem that you are in a position to accomplish a great deal of good.

Unfortunately, Archimedes's chronophone comes with certain restrictions upon its use:  It cannot transmit information that is, in a certain sense, "too anachronistic".

You cannot suggest, for example, that women should have the vote.  Maybe you could persuade Archimedes of Syracuse of the issue, and maybe not; but it is a moot point, the chronophone will not transmit the advice.  Or rather, it will transmit the advice, but it will come out as:  "Install a tyrant of great personal virtue, such as Hiero II, under whose rule Syracuse experienced fifty years of peace and prosperity."  That's how the chronophone avoids transmitting overly anachronistic information - it transmits cognitive strategies rather than words.  If you follow the policy of "Check my brain's memory to see what my contemporary culture recommends as a wise form of political organization", what comes out of the chronophone is the result of Archimedes following the same policy of looking up in his brain what his era lauds as a wise form of political organization.

You might think the next step would be to prepare a careful series of Plato-style philosophical arguments, starting from known territory, and intended to convince an impartial audience, with which to persuade Archimedes that all sentient beings should be equal before the law.  Unfortunately, if you try this, what comes out on Archimedes's end is a careful series of Plato-style philosophical analogies which argue that wealthy male landowners should have special privileges.  You followed the policy of "Come up with a line of philosophical argument intended to persuade a neutral observer to my own era's point of view on political privilege," so what comes out of the chronophone is what Archimedes would think up if he followed the same cognitive strategy.

In Archimedes's time, slavery was thought right and proper; in our time, it is held an abomination.  If, today, you need to argue that slavery is bad, you can invent all sorts of moral arguments which lead to that conclusion - all sorts of justifications leap readily to mind.  If you could talk to Archimedes of Syracuse directly, you might even be able to persuade him to your viewpoint (or not).  But the really odd thing is that, at some point in time, someone must have turned against slavery - gone from pro-slavery to anti-slavery - even though they didn't start out wanting to persuade themselves against slavery.  By the time someone gets to the point of wanting to construct persuasive anti-slavery arguments, they must have already turned against slavery.  If you know your desired moral destination, you are already there.  Thus, that particular cognitive strategy - searching for ways to persuade people against slavery - can't explain how we got here from there, how Western culture went from pro-slavery to anti-slavery.

The chronophone, to prevent paradox, will not transmit arguments that you constructed already knowing the desired destination.  And because this is a law of physics governing time travel, the chronophone cannot be fooled.  No matter how cleverly you construct your neutral-sounding philosophical argument, the chronophone "knows" you started with the desired conclusion already in mind.

The same dilemma applies to scientific issues. if you say "The Earth circles the Sun" it comes out of the chronophone as "The Sun circles the Earth". It doesn't matter that our civilization is right and their civilization is wrong - the chronophone takes no notice of facts, only beliefs and cognitive strategies. You tried to transmit your own belief about heavenly mechanics, so it comes out as Archimedes's belief about heavenly mechanics.

Obviously, what you need to transmit is the scientific method - that's how our own civilization went from geocentrism to heliocentrism without having the destination already in mind. Unfortunately, you also can't say to Archimedes, "Use mathematical laws instead of heroic mythology to explain empirical phenomena." It will come out as "If anyone should throw back his head and learn something by staring at the varied patterns on a ceiling, apparently you would think that he was contemplating with his reason, when he was only staring with his eyes... I cannot but believe that no study makes the soul look on high except that which is concerned with real being and the unseen." (Plato, The Republic, Book VII.) That is Archimedes's culture's stance on epistemology, just as science is your own culture's stance.

Can you suggest that Archimedes pay attention to facts, and authorities, and think about which one should ought to take precedence - by way of leading him down a garden path to the scientific method? But humanity did not invent the scientific method by setting out to invent the scientific method - by looking for a garden path that would lead to the scientific method. If you know your desired destination, you are already there. And no matter how you try to prevent your garden path from looking like a garden path, the laws of time travel know the difference.

So what can you say into the chronophone?

Suppose that, at some point in your life, you've genuinely thought that the scientific method might not be correct - that our culture's preferred method of factual investigation might be flawed. Then, perhaps, you could talk into the chronophone about how you've doubted that the scientific method as commonly practiced is correct, and it would come out of the chronophone as doubts about whether deference to authority is correct. After all, something like that must be how humanity got to science from nonscience - individuals who genuinely questioned whether their own culture's preferred method of epistemological investigation was correct.

If you try to follow this strategy, your own doubts had better be genuine. Otherwise what will come out of the chronophone is a line of Socratic questioning that argues for deference to authority. If your doubts are genuine, surface doubts will come out as surface doubts, deep doubts as deep doubts. The chronophone always knows how much you really doubted, and how much you merely tried to convince yourself you doubted so that you could say it into the chronophone. Such is the unavoidable physics of time travel.

Now... what advice do you give to Archimedes, and how do you say it into the chronophone?

Addendum:  A basic principle of the chronophone is that to get nonobvious output, you need nonobvious input.  If you say something that is considered obvious in your home culture, it comes out of the chronophone as something that is considered obvious in Archimedes's culture.

Comments (78)

Sort By: Old
Comment author: Joseph_Hertzlinger 23 March 2007 06:39:23PM 7 points [-]

Some of Archimedes most potentially-important research involved things he regarded as trivial toys. So if we advise him to get interested in Rubik's cube...

Comment author: Eliezer_Yudkowsky 23 March 2007 06:50:26PM 0 points [-]

Joseph, the chronophone won't transmit a literal description of a Rubik's Cube any more than it will transmit a description of a printing press. But if you've ever mused over apparently trivial toys, you could talk about your own musings into the chronophone, and it would be translated appropriately on the other end. What appropriate "trivial toys" have you mused over?

Comment author: ESRogs 05 April 2013 07:11:06PM 3 points [-]

Wasn't Joseph already using the Rubik's cube as an example of a trivial toy?

Comment author: pure-awesome 04 May 2013 11:45:02PM 6 points [-]

Yes, I don't think Joseph's intention was to get Archimedes to understand a Rubix cube. I believe his intention was to get Archimedes to play with 'trivial toys' and so he thought talking about Rubix cubes might do the trick.

Comment author: Mike_Linksvayer 23 March 2007 07:16:35PM 3 points [-]

I would attempt to send useful facts that the chronophone should not distort -- facts that classical Greek culture was ignorant of, but not biased against -- discoveries that would have been accepted by Greek culture, had they been made. I'd have to think about what those would be.

Comment author: Mike_Linksvayer 23 March 2007 07:24:26PM 2 points [-]

The sort of thing I have in mind is mathematical proofs and engineering designs that go just beyond what he was able to manage without my help, not imparting the scientific method or liberal morality. Given what the chronophone does more ambitious communication seems really risky.

Comment author: Robin_Hanson2 23 March 2007 08:19:52PM 8 points [-]

It would sure help if you could speak into one end and then hear how it comes out on the other end, and then edit and repeat. If so, I'd try to encourage him to try to make useful devices that make money, and to create a tradition of this activity.

Comment author: rcriii2 23 March 2007 08:59:20PM 3 points [-]

It sounds like if Archimedes asked us a question, he would almost always get the answer he was expecting. Given this, would he be better off calling someone closer to his own time, whose answers were less likely to be 'too anachronistic', and thus more likely to give him new information? Maybe our best bet would be to refer him to, say, James Madison or Galileo.

Comment author: qer 23 March 2007 10:15:14PM -1 points [-]

Teach him decimal notation! It would bring us ahead a thousand years. Can't remember who first said that, but I read it somewhere.

Comment author: themusicgod1 31 January 2013 12:04:08PM 0 points [-]

Wouldn't he have just discarded it as he was trained with other notation?

It would be like someone in the modern english world trying to learn chinese math notation. We could probably understand the concepts so could in principle do it, but it would seem relatively unweildly, even if it turns out that it's a much more elegant way of doing math. We'd never know.

He might very well learn it and then go "that's cute" and then ignore it.

Comment author: Tom3 23 March 2007 10:44:22PM 13 points [-]

I would make an argument for the importance of space exploration and colonisation, on the assumption that in Archimedes' time this would be heard as an argument in favour of naval expansion. Since the greeks were seafaring people anyway, Archimedes might be convinced to push for a great greek colonisation program. Who knows? They could reach the far east, or even the americas. The age of global trade might begin two thousand years earlier.

Comment author: Raw_Power 31 October 2010 04:54:21PM 4 points [-]

Alexander tried this by land. It didn't work out, but it did leave us Hellenism.

Comment author: Eliezer_Yudkowsky 23 March 2007 11:14:42PM 7 points [-]

So far it looks like Tom is the only one who made a non-meta valid suggestion. Come on, surely we can do better? For example, I could speak in my concerns about "statistical significance" levels, hoping it would come out as some kind of criticism of Greek philosophers' criteria of argument.

Comment author: Eliezer_Yudkowsky 23 March 2007 11:30:13PM 1 point [-]

Robin, suppose you can't fine-tune the chronophone that way. How would you implement your object-level suggestion - what would you speak into the chronophone? Remember, a special-purpose garden path won't work. So, what cognitive efforts have you put forth to make useful devices that make money?

Comment author: HalFinney 23 March 2007 11:36:23PM 12 points [-]

It's a great thought experiment, but a little confusing. I'd suggest that a better name for this is the "metaphone", in that whatever you say it maps to the analogous concept in the recipient's mind and worldview. It's an interesting concept even beyond the specifics of time travel. I don't know if it could really work though, as it is too hard to figure out what would correspond in people's minds, in enough detail.

There's also a question of how meta it goes. Tom's idea is a good one if it just maps space exploration to naval exploration (assuming that that is in fact something we want to encourage). But maybe the phone would go meta even on the concept of exploration, and map that concept to something more analogous in Archimedes' world view. We come from a culture which is based on centuries of success via exploration, so maybe the phone would map that to something that had provided centuries of success for the greeks, like philosophy or something.

Comment author: Mike_Linksvayer 24 March 2007 12:47:07AM -2 points [-]

Tom, the Greeks were colonizers, e.g., Syracuse is not in the homeland. http://en.wikipedia.org/wiki/Colonies_in_antiquity#Greek_colonies_.28.22apoikiai.22.29

The obvious non-meta version of my suggestion is to teach Archimedes calculus.

Comment author: Peter_de_Blanc 24 March 2007 01:21:27AM 19 points [-]

You should ask the greatest mathematician of the ancient world to work on FAI theory. If he solves the analogous problem, then when he explains his solution to you over the Chronophone, it'll come out on your end as a design for an AI.

Comment author: MathieuRoy 02 February 2014 06:40:00AM 4 points [-]

Nice :-) but in fact, the chronophone will transmit a problem just as hard for Archimede as FAI is to us. So he'll probably solve the problem in the same amount of time than us (so it won't help us). I wonder what would be this problem? Is going do the Moon for Archimede just as hard than building a FAI for us?

Comment author: Strange7 06 April 2014 01:35:58AM 1 point [-]

In that case, Archimedes and his successors spend a few hundred years working on the problem, and then recalibrate the chronophone to communicate over a correspondingly shorter interval, just so they can seem smarter by giving the answer immediately.

Comment author: Eliezer_Yudkowsky 24 March 2007 01:56:05AM 4 points [-]

Ger, the chronophone won't directly transmit decimal notation - it will come out as Roman Numerals. Maybe you could get a computer programmer who works on simple user interfaces, or a mathematician trying to teach category theory to high-school students.

Linksvayer, if you speak calculus into the chronophone, it comes out as some form of mathematics that is widely known in Archimedes's era - maybe Pythagoras's musical harmonics. Perhaps Andrew J. Wiles could speak into the chronophone.

de Blanc, for obvious reasons, Archimedes communicates literally with us if he can communicate at all - he just writes down his reply somewhere and hopes that it gets copied through a few thousand years. Transmitting information from the past to the future is conceptually straightforward.

Hal, you indeed seem to comprehend what I was aiming at with the chronophone parable. I may even have responded to Hanson incorrectly. In our time it is already widely believed, at least among the educational elite, that making money (especially by inventing new gadgets) benefits humanity. So if you said that into the chronophone, it might come out as a paean to the benefits of conquering enemy territories. In Archimedes's time a technological marketplace is a nonobvious idea; and a basic principle of the chronophone is that to make something nonobvious come out in the past, you have to say something nonobvious in the present.

Comment author: Mike_Linksvayer 24 March 2007 06:09:34AM 12 points [-]

Archimedes lived two centuries after Pythagoras and knew more math in spite of being known as an engineer. http://en.wikipedia.org/wiki/Archimedes#Mathematics

However, calculus is obvious now, so that won't work.

But the task is now easy -- I'd talk to Archimedes about plant rights (see http://www.overcomingbias.com/2007/03/morality_of_the.html).

Comment author: JenniferRM 12 June 2012 07:10:49PM 3 points [-]

Interesting. I think this comment could plausibly be the primary inspiration for HP:MoR's chapter 48 :-)

Comment author: michael_vassar 24 March 2007 07:55:13AM 4 points [-]

And if I talk to Archimedes about Greek-Style philosophy or mathematics? Does this come across as discussion of Egyptian philosophy and mathematics or does it remain Greek? I can find all sorts of interesting proposals in the latter, including governmental forms etc that sound promising by todays standards as well as a lot of fairly good science from the pre-Socratics.

Comment author: Matthew_Pianalto 24 March 2007 03:55:13PM 1 point [-]

"Archimedes, don't listen to what Aristotle said about Empedocles!" (Empedocles proposed a rough version of natural selection, and Aristotle rejected Empedocles' view as absurd in his Physics.) I suppose that might be "too anachronistic," depending upon just how much authority was given to Aristotle's teleological view of the world. I don't know that saying this would be forcing our own preferred conclusions on him.

Comment author: LuigiG 24 March 2007 05:15:16PM -2 points [-]

I'd put Archimedes through to somebody like the current president of the United States ("Dubya") or maybe some religious fundamentalist. Archimedes would be so disappointed and perhaps horrified about what kind of people are hanging around in the future that he may try to use his intelligence to figure out what has gone (or will go) wrong in history and do something about it.

Comment author: Nick_Bostrom2 24 March 2007 08:01:08PM 6 points [-]

I'm not sure I understand exactly how the chronophone works. It sounds a bit like the only useful ideas a person can transmit are ideas that she herself has independently worked out or discovered; in which case not the same ideas but some analogous and similarly useful ideas gets delivered to Archimedes. In this case, I guess I might try to read out some of my research papers, hoping that they contain some useful original insights. It might also work if I transmit ideas that have originated with others but whose merits I have grasped through my own independent judgement.

It seems if you subtract all the information advantages that we moderns have, all that remains in this exercise are the organic qualities of our brains and the amount and quality of intellectual labour that our brains have performed.

Comment author: Bill_Kaplan 24 March 2007 08:19:44PM 0 points [-]

Transmit: "All things are made of atoms." Democrates said it. Confirm it.

About politics, do the same. Confirm Democrates.

Comment author: Anna 25 March 2007 01:11:30AM 0 points [-]

Suppose you could send messages back in time to Archimedes of Syracuse, using a chronophone which - to avoid transmitting anachronistic information - transmits the results of executing cognitive strategies, rather than words.

"You have no idea what you're about to discover!" "What you thought is real, is really not the truth!" "You can't imagine the facts!" "What is real, is only real to you!"

Just some ideas. Anna

Comment author: jsabotta 25 March 2007 03:15:58PM 9 points [-]

Really, the best advice would be "Get out of Syracuse before the Romans show up" but I suppose your infuriating machine would refuse to transmit that as well.

Comment author: goldfishlaser 04 August 2009 12:22:27AM 1 point [-]

Hey, all you have to do is think of an analogous suggestion today. I'm afraid I lack the historical knowledge necessary to even attempt to provide direct advice...

A possible problem would that there would be no analogous situation. That there is no place for your country to get out of before some people show up. I don't know if communicating similarly important abstract ideas works, and how much your correctness or incorrectness of advice matters...

Comment author: taryneast 31 May 2011 08:28:12AM *  1 point [-]

A possible problem would that there would be no analogous situation.

Get out of Austria before the Nazis show up.

Comment author: johnlawrenceaspden 08 October 2012 05:16:50PM 1 point [-]

But that's hindsight, so it would come out as advice about Archimedes' past.

Comment author: taryneast 04 November 2012 11:58:17PM 1 point [-]

It's the use of a past example to advise on future actions.

Comment author: PrimIntelekt 10 October 2009 05:22:32PM 5 points [-]

Perhaps "Get off of Earth before Nemesis sends some more comets our way"

Nemesis: http://en.wikipedia.org/wiki/Nemesis_%28star%29

Reasoning: The Roman invasion could have been predicted, but wouldn't have been well supported by evidence -- although foreign invasions were certainly a repeated (even cyclic) phenomenon. Of course, the Nemesis hypothesis rests on an unproven root cause of all major extinction events, whereas invasions came from various, known causes...

Comment author: Pavel 26 March 2007 04:35:17PM 0 points [-]

What about e.g. "Time is a means of measuring movement"?

Comment author: _Gi 26 March 2007 06:04:58PM 1 point [-]

What happens when a contemporary political and economical analysis is sent over, including the troop sizes of contemporary armies, supply levels, economic conditions of the surrounding countries, facts that could not be available to Archimedes but could be available to some of his contemporaries?

Comment author: michael_vassar 27 March 2007 08:28:47PM 2 points [-]

Jsabotta: Easy to transmit that one. Get out of NYC, DC, and to a lesser degree other large US cities before terrorists nuke it/them. Maybe I should do that. OTOH, there are practical reasons to be there as well.

All I can say is "I'm working on it".

Comment author: Aaron_Davies 28 March 2007 04:27:36PM 0 points [-]

I'd start by looking into how the scientific method was developed in our history, and seeing if that would be any help. Perhaps sending him some of Roger Bacon's work would be useful?

Comment author: RobinZ 17 February 2010 08:37:37PM *  13 points [-]

Rereading this article from Emile's link:

I think my major problem with this article is that the perfectly reasonable conclusion - that you can't do better in the future today by thinking today's cached thoughts about how people in the past could have done better in the future of the past - is obscured by the utterly ridiculous device of the eponymous chronophone.

To elaborate: If you analyze the question, "how did people in the past figure out that ideas are tested by experiment?", then you can immediately rule out "by asking how people at their own time evaluated ideas". And indeed, generalizing to "how did people in the past become smarter than their contemporaries - i.e. better at solving their problems?", you see that it is trivially true that you can't count on the standardized thinking of the present to take you beyond the standardized thinking of the present.

But if you analyze the question, "what do you say into the chronophone to convince Archimedes?", you come up with "this thing couldn't possibly work - there is no way to draw a unique relation between rationalizations of current and past ideas, so it fails sci-fi". Which has nothing to do with anything.

Comment author: Mass_Driver 10 May 2010 07:13:49AM 2 points [-]

I've always thought that the middle chapters of Plato's Republic were satire. Are you quite sure that classical Greek philosophers were disdainful of empirical phenomena? How do you explain, e.g., Aristotle's Physics?

Comment author: Jack 10 May 2010 08:45:41AM *  1 point [-]

I've always thought that the middle chapters of Plato's Republic were satire.

Nope. Or at least that is an extremely novel interpretation, as far as I know.

Are you quite sure that classical Greek philosophers were disdainful of empirical phenomena? How do you explain, e.g., Aristotle's Physics?

Also, all his biology. I agree the post is unfair to the ancient Greeks. Aristotle invented observational science. The problem is he didn't think to run any experiments. He just collected stamps, so to speak.

Comment author: rebellionkid 09 January 2012 05:35:51PM 2 points [-]

My understanding is not that he "didn't think to" run experiments, rather he actively rejected the doing of experiments. The idea is that one can study nature, study things as they naturally are, because that is them acting in accordance with their inherent properties. But it makes no sense to study things in a lab, in an artificial environment. If I launch something out of a cannon then I disturb it from how it inherently acts. Of course something acting against how it inherently acts is pointless to study. Likewise doing any kind of experiment will only tell you about your experiment (at best) and tell you nothing of the natural state of things.

Comment author: thomblake 09 January 2012 06:07:28PM 1 point [-]

I've always thought that the middle chapters of Plato's Republic were satire.

I've held the same view, and while it is unpopular it's not unique.

Comment author: TheOtherDave 22 October 2010 04:15:50PM 12 points [-]

The honest answer to your question is "Beats the !@#!$ out of me."

But if abdicating to someone smarter isn't an option, I need to do something. And starting from my values is asking for trouble: if they are temporally embedded the chronophone won't transmit them; if they aren't Archimedes has just as much access to them as I do.

So I have to start from the other side: model Archimedes, decide what modifications I want to introduce to that model, work out what aspects of his own mind will induce those modifications if excited/inhibited sufficiently, work out what aspects of my mind are most analogous to those, identify ideas that excite/inhibit my corresponding cognitive structures, and communicate those ideas.

So far, so good... this is essentially the same way I influence anyone in the real world who doesn't already happen to agree with me. It's also more or less how I train my dog. "Seek first to understand, then to be understood."

(Were this a real situation, I'd start by doing a lot of research on Archimedes and his milieu, about which I'm pretty ignorant. I'm not going to do that here, so my suggestions will be pretty bogus.)

Suppose I decide to influence Archimedes towards social equality (e.g., eliminate slavery, wider voting rights, that sort of thing).

Let's assume his current opposition to this idea is in part going along with social norms, in part the notion that there really are fundamental differences between people that map to their social roles, and in part the practical acknowledgment that restricting power to his own class makes his life better.

I can't do much about the first or the third. He could teach me a thing or two about being willing to stick your neck out, and I agree that there are practical benefits to preserving privilege; I could hardly convince him otherwise given the rhetoric-flattening nature of the chronophone. But the second seems like a potential lever.

There's no way to convey my own belief that the differences aren't as fundamental as all that, nor any reason Archimedes would be convinced by it. So I guess I'd start by thinking about experiments that have led me to reject intuitively obvious beliefs I've had about cognition. Demonstrations of confirmation bias, of conformity effects, of the use of availability heuristics in decision-making, that sort of thing.

The specifics don't matter much; what I'd want to get across is the idea that people, and how people interact, and what people are capable of, is just as subject to careful investigation without distortion by prior expectation as, say, machinery or the density of crowns.

If I could do that, and thereby encourage him to begin empirically exploring cognition... that'd be cool. A 2000-year headstart on that could really change the world.

I don't think much of my chances, though. I've tried to do this too many times across a much smaller gulf with better tools and failed miserably.

Comment author: nick012000 22 October 2010 07:05:24PM 0 points [-]

It's simple. I'd make the best damn argument for slavery I could, knowing that the chronophone will invert it into the best damn argument against slavery I could give.

Comment author: tenshiko 25 October 2010 02:51:24AM 5 points [-]

If I'm understanding the chronophone correctly, the thing is that what comes out cannot be anachronistic. Maybe I'm not. If I'm understanding it as a strategy-conveying phone, then it would just tell Archimedes that you're trying to trick him into believing an anachronism.

Personally I'd eagerly chirp about awesome technology like nanobots and solar power and high-speed trains that so many countries seem to be ignoring right now, and hope that it would pick up the kind of stuff Hero of Alexandria was doing with steam and such. (Although this might not work. It's such a shame that he came after Archimedes). It would be very interesting to see how this would turn out in conjunction with the whole space travel --> naval expansion idea...

Comment author: nick012000 25 October 2010 08:51:41AM 1 point [-]

No, it'd tell him that you're arguing against the local belief structure regarding slavery. In his time, it'd be an argument against slavery.

Comment author: VAuroch 14 January 2014 02:37:12AM 3 points [-]

It seems more likely to come out as an argument for some practice that had been given up as immoral in his time. i.e. human sacrifice, which I believe was no longer part of Greek religion by that time.

Comment author: moshez 27 April 2011 06:55:24PM *  16 points [-]

I know it's ancient, but what the hell, I finally have answers!

"Archimedes, you should teach everybody to program. Every boy, every girl, at as early an age as you find possible. Computers are all around us, and knowing how to program gives us power over them, and lets us be in control of our universe."

I hope it comes out as

"Archimedes, you should teach everybody to read, write and the basic of mathematics. Every boy of reasonable upbringing, at as early an age as you find possible. Scrolls full of words and math are all around us, and knowing how to parse them give us power over them, and lets us be in control of our universe."

I'm kind of sad that it comes out as "every boy of reasonable upbringing", but since I see "every boy, every girl" as executing my time's algorithm of "who can learn" (I didn't propose teaching chimps or dolphins), that's the best I can do. Regardless, if you think how programming is treated as a specialization now, instead of as a basic art, I hope that I actually conveyed something useful to Archimedes.

Also: "Archimedes, copyright and DRM are evil. They hamper the transfer of knowledge, and make people get used to being slaves to others' whims."

If I'm very, very lucky, that might come out as an argument against slavery. If I convince Archimedes to end slavery, I consider it a big big win :)

Comment author: Daniel_Burfoot 17 May 2011 04:17:20AM 9 points [-]

I'm kind of sad that it comes out as "every boy of reasonable upbringing", but since I see "every boy, every girl" as executing my time's algorithm of "who can learn" (I didn't propose teaching chimps or dolphins), that's the best I can do.

!! This obviously implies that we should be trying to teach chimps and dolphins!

Comment author: moshez 18 May 2011 11:00:51PM *  3 points [-]

That does not follow. I followed the algorithm of "check who my counter-parts think is teachable according to the most cutting edge research", so Archimedes follows the same algorithm. However, I happen to really truly believe that my time's beliefs on this specific issue are fairly accurate -- and that even though Archimedes probably believed the same about his time, he's wrong and I'm right.

Comment author: AdeleneDawner 19 May 2011 12:03:00AM 3 points [-]

he's wrong and I'm wrong.


Comment author: moshez 19 May 2011 11:05:13PM 0 points [-]

Yes :)

Comment author: Clippy 19 May 2011 12:32:31AM 2 points [-]

Yes, humans should teach programming to chimps and dolphins. Who but a racist would disagree?

That doesn't mean that you must teach them that specific skill now, with their current knowledge -- there are pre-requisites you will have to cover first. But you certainly need to attempt the first steps.

Comment author: FAWS 19 May 2011 11:21:22PM *  2 points [-]

Well, some people are trying to teach chimps sign language or similar. Success there seems likely to be a necessary prerequisite for teaching them programming.

Comment author: taryneast 31 May 2011 08:24:30AM 0 points [-]

I would speak highly of sharing scientific research with China and India - in the hopes that it would come out as a paean on the benefits of sharing research with other cultures - hopefully leading to a sharing of information with the great Islamic scientists. Most of which are after Archimedes' time... but having a culture-of-sharing in place would be a good way to spark things off.

Comment author: rebellionkid 09 January 2012 05:44:47PM 1 point [-]

Archimedes died in 212 BC. Justinian closed the pagan schools and killed of the classical research project in 529 AD. The first book of the Quran was delivered in 610 AD. So sharing with the Islamic scientists (who had all read Aristotle anyway) doesn't make sense.

But an expansion of the communication between places like Alexandria, Athens, Cos etc could be a really good plan.

Comment author: orthonormal 20 June 2011 10:15:54PM 3 points [-]

Another way of putting the idea here: What ideas could qualify as X, where X is to modern common sense as modern common sense is to ancient Greek common sense?

Comment author: khafra 21 June 2011 02:02:09PM 0 points [-]

The idea here might also be: "Which memes M form a group acted on by a torsor of cultural zeitgeists C such that MC1=C2, where C2 is a more desirable cultural zeitgeist than C1?"

I'm honestly not sure which meaning Eliezer originally intended, as both seem valuable focuses for contemplation.

Comment author: Benquo 21 June 2011 02:19:52PM 0 points [-]

Don't worry if the simplest mathematical description of the appearances defies common sense. Common sense will catch up. Now rethink astronomy and pay attention to interesting regularities.

If something counts as "evidence" in one domain but not in another, you should think long and hard about whether this different treatment is justified. Maybe you're privileging "evidence" that should be ignored, or throwing out valuable information.

Comment author: Benquo 21 June 2011 02:21:33PM 1 point [-]

Which sciences are thought to be noble now is not a perfect indicator of which ones will prove to be valuable in the future.

Comment author: EvelynM 27 June 2011 04:44:55PM 3 points [-]

Horoscope: Imagine your future self telling you an important idea, which you can put to use right now. What would that idea be?

Comment author: sniderj1 07 October 2011 04:45:16AM 2 points [-]

The first thing I'd do is convince him that I'm a messenger from a God. So, I'd say "If I had a super-scientific machine that gave me all the answers, I'd put it on a big pedestal in George Mason's engineering building and use it to proofread everyone's research."

And he'd hear "If I had a mystical way of talking to the gods, I'd put in on a big pedestal in the agora and have philosophers talk to it all day."

Once that happens, whatever we want to say will be much more convincing.

Comment author: potato 10 December 2011 09:38:51PM *  0 points [-]

What if I talked to him about pebbles and buckets? Did the Greeks have pebbles and buckets? Did they have a word for probable? Or a way to express it? Cause then i would just try to get them to get bayes theorem, and priors, and updating, and bayes networks, through talking about pebbles with different properties in buckets of different sizes. And the whole time say that this is just like belief. Or that in the future, this is epistemology.

I would probably try to explain a bit of Boole to them first, use black and white pebbles. Just conjunction negation, and disjunction. Show them demorgan's.

I think this would all work as long as i give examples in terms of plain things that were around back then, and let them do the generalizing.

I would hope that if i showed this to the right cat back then, I might inspire some early empiricism.

Comment author: Insert_Idionym_Here 30 January 2012 09:49:09PM 0 points [-]

It seems that in order to get Archimedes to make a discovery that won't be widely accepted for hundreds of years, you yourself have to make a discovery that won't be widely accepted for hundreds of years; you have to be just as far in the dark as you want Archimedes to be. So talking about plant rights would probably produce something useful on the other end, but only if what you say is honestly new and difficult to think about. If I wanted Archimedes to discover Bayes' theorem, I would need to put someone on the line who is doing mathematics that is hundreds of years ahead of their time, and hope they have a break-through.

Comment author: johnlawrenceaspden 08 October 2012 05:28:30PM 3 points [-]

I think probability theory would have been very accessible to the Greeks, had they only thought to think about games of chance, which they certainly played. I bet if you'd asked Archimedes 'What odds should you offer on a bet that two dice get seven?', then the whole thing would have come crashing out within a hundred years or so.

So you might want to put him in touch with a modern philosopher trying to take a mathematical approach to something mysterious, say Dennett.

Comment author: keddaw 24 July 2012 01:16:50PM 0 points [-]

I think I have a couple of suggestions:

  1. Find a current social/scientific norm that maps almost perfectly to Archimedes' time that you disagree with (e.g. the need for a strong, expansionist military; use of torture; increase in state power over citizens; existence of a political class etc.) and use the same arguments against it now, including outcomes and dangers, and they should map to decent arguments then.

  2. Find a socially wronged group and use arguments for their emancipation (women, non-whites, gays, children etc. depending on your era) and whichever hated/discriminated against group they map to in Archimedes' time will almost definitely be helped by your emancipation arguments.

  3. Say that hatred/fear based on religion is wrong, but that religion itself is wrong and the acquiescence to authority is also problematic.

  4. Describe the problems of current economic bubbles, their causes and possible solutions to avoid them. This will hopefully help with the basic economy of ancient Greece.

  5. Argue for non-objective morality. And moral error theory. And a lack of libertarian free will. And against fate (but for a QM-accepting variant of determinism).

Basically, as long as your argument is against the common knowledge/understanding of your time you have a chance of getting across a decent version of it, or the principle at least, especially if the common knowledge has a (close) corollary in Archimedes' time. (HT. unequally-yoked).

Comment author: MixedNuts 24 July 2012 03:06:38PM 3 points [-]

No, it'll come out as arguments for then-controversial position that Archimedes already holds.

Comment author: Philip_W 23 September 2012 07:20:10PM 0 points [-]

The problem with the chronophone is that almost all of our knowledge and beliefs comes from the authority of others. Unless you have designed every stage of the test yourself, and then observed that no errors were made and if you have derived all the underlying assumptions yourself (or have them in common with Archimedes), you can't use the result of such a test because you've accepted the authority of the testers.

And still you could argue that verifying the accuracy of every test you have ever done is merely applying Bayesian Reasoning/The Scientific Method as it was taught you - which would come out as applying Aristotelean philosophy, or something close. Tell him to educate all children of the world? You can't even tell him to do a damned scientific experiment!

If we want to convince Archimedes of anything, we're going to have to rederive science and ethics from principles which Archimedes would agree with for the same reasons as us (or, to generalize to the presumable meaning of the post, principles which are actually fundamental).

Comment author: johnlawrenceaspden 08 October 2012 05:13:56PM 0 points [-]

It seems to me that the two most scary risks we face are nanotechnology and AI, and that compared to them, nothing else matters at all. So it seems like our best strategy would probably be to try to understand those risks as thoroughly as possible, and to throw vast amounts of money at research to see whether there's any way to deal with them safely.

So if that's our best plan, and I say it into the magic time idea translating phone, maybe something good will come out of Archimedes' end, like 'try to understand the Roman Republic and how you might counter it before it shreds you'. And who knows. Maybe if the Greeks and the Carthaginians had understood the nature of the threat, they might have been able to do something about it.

But it's not clear that preventing the rise of the Roman Empire would optimize the modern world.

Comment author: themusicgod1 31 January 2013 02:23:42PM *  1 point [-]

I would explain the concepts of the craziest, most non-obvious ideally moral ideas I've ever had [such as the idea that Nick Bostrom's Interstellar Opportunity Cost paper completely changes the nature of the pro-life debate, such that it no longer is sensible to freeze all fetuses instead of aborting them, instead, if we're serious about being pro-life we should crop humanity down to only what is needed to spread human life to other stars, and that there are economic considerations to freedom but that they are subtle and complex]. Something that is so ...off the wall might not go through directly, but it might come through as something equally out there [that the greeks should dedicate all their energies and efforts to seafaring and trade]

And in fact the further in the future I would be aiming to talk about the better. What matters about 2,000 paltry years when we're talking about post-singularity or near-singularity times? The differences between us will be minor in comparison, and not be 'smudged' by the chronophone.

Comment author: Manfred 31 January 2013 03:07:35PM 7 points [-]

Sadly, I'm afraid all these suggestions were generated with the cognitive algorithm "think of something that will make the chronophone output something good," and will thus result in arguments that Archimedes would have come up with when trying to get the chronophone to output something good.

Comment author: sjmp 20 April 2013 03:59:04PM 1 point [-]

I suppose I could talk about how I've never had any other proof about existence of photons except what I've read in books and I've been told by teachers. How I am in fact taking science on authority rather than first hand experimental proof. Sure, scientific method and all the explanations sound like they make sense, but is that enough to accept them as facts? Or should I lower my probabilities until I actually find out myself?

Comment author: smijer 17 August 2013 12:13:21PM 3 points [-]

70 comments so far, and none of them, "Holy Shit! I'm talking to Archimedes!"

... which I suppose he would hear as "Ye Gods! I'm talking to Plato"...

Comment author: MathieuRoy 02 February 2014 07:20:31AM *  0 points [-]

If doubting is/was accepted in our current society, and we wanted Archimede to doubt about his beliefs, would we have to doubt about the value doubting, or be certain about the value of doubting?

It's a joke. As Eliezer said "to get nonobvious output, you need nonobvious input", so obviously, we'd just have to find something nonobvious. :-)

I wonder if we will ever come up with something that is as nonobvious to us right now as bayesian thinking was to Archimede.

Comment author: Ender 19 April 2014 05:13:01AM 0 points [-]

You might pull together a good message just based on the original question, "what advice do you give to Archimedes, and how do you say it into the chronophone." Yudkowsky's question was designed to make us think non-obvious thoughts, after all.

"Would you be able to ask anything meaningful through the chronophone?"

(My construction might not be quite right. I'm feeling all smug and Godelian, but it's 1 AM, so I've probably missed something.)

Comment author: Jiro 19 April 2014 10:52:41AM *  1 point [-]

The description of what the chronophone transmits is vague enough that no matter what I say, there's always a plausible way to interpret that description such that it would fail to work. As you, as the person posing the problem in the first place, get to decide which interpretation is correct, that means there's no solution.

By the time someone gets to the point of wanting to construct persuasive anti-slavery arguments, they must have already turned against slavery. If you know your desired moral destination, you are already there. Thus, that particular cognitive strategy - searching for ways to persuade people against slavery - can't explain how we got here from there, how Western culture went from pro-slavery to anti-slavery.

I don't buy this argument either. Someone might not want to specifically construct persuasive arguments against X. Rather, he tries to construct persuasive arguments about X regardless of they are for or against it. Then he decides to accept or reject X depending on whether the persuasive argument he created happened to be for or against it (possibly changing his preexisting opinion about X in the process). Once he has constructed the persuasive argument (either for or against) and has his newest opinion for or against X, he then tries to convince others of it.

It also ignores the difference between an argument being persuasive to yourself and to others. Perhaps some believers in X and some believers in ~X want to come up with arguments that confirm their preexisting position. Both of them manage to produce arguments that convince themselves. But it turns out that when they use those arguments on other people, they're not equally effective in convincing other people. In fact, it turns out that believers in X can convince a subset of believers in ~X to switch sides, while believers in ~X cannot convince a similar subset of believers in X to switch sides.

(And it might be hard to see this because once the arguments have spread, when you get to the next generation, the arguments are already floating around, so the people who if they had been born earlier would have started out believing in ~X but gotten converted, would have started out already believing X and had no need to change their mind. You would then observe this and think "nobody can be convinced by argument to change their mind" when in fact all the people who are susceptible to persuasion by argument are already on the side that has the persuasive arguments.)

Comment author: ike 18 June 2014 09:52:18PM *  1 point [-]

Ger suggested teaching Archimedes decimal notation. Well, if you speak decimal notation - our home culture's standard representation of numbers - into the chronophone, then the chronophone outputs the standard representation of numbers used in Syracuse. To get a culturally nonobvious output, you need a culturally nonobvious input. Place notation is revolutionary because it makes it easier for ordinary people, not just trained accountants, to manipulate large numbers. Maybe an equivalent new idea in our own era would be Python, which makes it easier for novices to program computers - or a mathematician trying to standardize on category theory instead of set theory as a foundation for mathematics.

– (From the follow up article)

I think that the rules for the chronophone are either inconsistent, or incorrectly applied. There is not a fine enough distinction between our “cognitive policies”, and our “meta-cognitive policies”.

That is, if you were to explain Python into the chronophone, you are executing two different cognitive policies at two different levels of reduction:

  • Explain a non-obvious kind of math in our culture which makes it easy for ordinary people to do things previously reserved for professionals. Or the meta-policy of
  • Say something into the chronophone which will result in decimal notation (something standard in our culture) coming out of the chronophone

Now, presumably, if Archimedes says something into the chronophone to himself, it comes out unchanged. Therefore, the rules predict that what will come out the chronophone will be (depending on which level of meta-cognitive-policy you choose):

  • A non-obvious kind of math in Archimedes's culture which makes it easy for ordinary people to do things previously reserved for professionals. E.g. decimal notation.
  • Something that Archimedes could say into the chronophone which will result in a standard representation of math in his culture coming out of the chronophone.

You see the problem here. If level 2 is the “correct” way of looking at it, (which seems more fundamental to me), then whatever you say, it has a root goal of getting Archimedes to understand something obvious in your culture. Even if you say something unobvious.

(I suppose you could say something “doubly-unobvious” to get it to say something merely unobvious in both his culture and ours, but it's unobvious to me what that would look like. What useful ideas are unobvious both to our culture and Archimedes, and more, are the same level of unobviousness, that you could iterate the unobviousness and get useful output? If that's too unclear, I can try to clarify it.)