« Share likelihood ratios, not posterior beliefs | Main | Our Biggest Surprise »

February 06, 2009


Eliezer, I didn't find your explicit arguments persuasive, nor even clear enough to be worth an explicit response. The fact that you yourself were persuaded of your conclusion by fiction does not raise my estimate of its quality. I don't think readers should much let down their guard against communication modes where sneaky persuasion is more feasible simply because the author has made some more explicit arguments elsewhere. I understand your temptation to use such means to persuade given that there are readers who have let down their guard. But I can only approve of that if I think your conclusions are worth such pushing.

Robin, it looks to me like we diverged at an earlier point in the argument than that. As far as I can tell, you're still working in something like a mode of moral realism/externalism (you asked whether the goodness of human values was "luck"). If this is the case, then the basic rules of argument I adopted will sound to you like mere appeals to intuition. I'm not sure what I could do about this - except, maybe, trying to rewrite and condense and simplify my writing on metaethics. It is not clear to me what other mode of argument you thought I could have adopted. So far as I know, trying to get people to see for themselves what their implicit rightness-function returns on various scenarios, is all there is and all there can possibly be.

Robin: Well, to be fair, it was written well enough that lots of us are arguing about who was actually right.

ie, several of us seem to be taking the side that the Normal Ending was the better one, at least compared to the True Ending.

So it's not as if we were manipulated into taking the position that Eliezer seemed to be advocating.

Also - explaining complex beliefs is a fantastically difficult enterprise. To consider fiction as just a way of "sneaking things past the unwary reader's guard" is selling it far short. It verges, I fear, on too much respect for respectability.

I tried to make the Superhappy position look as appealing as possible - show just how strange our position would look to someone who didn't have it, ask whether human children might be innocent victims, depict the human objections as overcomplicated departures from rationality. Of course I wanted my readers to feel Akon's helplessness. But to call that "sneaking things past someone's guard" is a little unfair to the possibilities of fiction as a vehicle for philosophy, I should think.

I'm still conflicted and worried about the ethics of writing fiction as a way of persuading people of anything, but conflict implies almost-balance; you don't seem to think that there's much in the way of benefit.

Wei Dai, Sure I too could invent numbers large enough to make any calculation give me the result I want. But as it still stands I think that 2^(10^20) is impractically huge fore any universe, especially one that seems to be based in our own. Also I think it's hard to imagine that the starline topography to be completely random among the whole universe. So I stand by my stance that it is difficult to imagine a universe where the super happies do not come across the humans again in a relatively short period of time. And I figure the point about fiction is that you're trying to convince you're readers about a consistent world where a story takes place in.

EY, but you are a moral realist (or at least a moral objectivist, which ought to refer to the same thing). There's a fact about what's right, just like there's a fact about what's prime or what's baby-eating. It's a fact about the universe, independent of what anyone has to say about it. If we were human' we'd be moral' realists talking about what's right'. ne?

Eliezer, academic philosophy offers exemplary formats and styles for low-sneak ways to argue about values.

Eliezers philosophy of fun and how it related to human value system was not grounded enough to be intelligble?

According which criterion is it balanced? How likely is it that extremely positive and strong feedback balances negative side-effect of manipulating people? Did you just justify use of fiction by feeling conflicted about it?
As a side note I'll comment that so far I have no reason to expect your story to be a excellent introduction to your ideas. It does show several ideas well, but readers not familiar with your writing miss a lot easily, but notice that you've used a lot of insider-speak. I have no reason to expect it to be terrible either.

Thom Blake:
Whatever is true of human rightness might not that much look anything invidual humans value, but it ought to address it all somehow. I'd like to hear if there is a reason to expect human rightness to be in some sense coherent and if there is then I'd like to understand in what sense there is. I don't remember from top of my head any posts addresing this.

Robin, what is your favorite piece of academic philosophy that argues about values?

Nicholas, our own universe may have an infinite volume, and it's only the speed of light that limits the size of the observable universe. Given that infinite universes are not considered implausible, and starlines are not considered implausible (at least as a fictional device), I find it surprising that you consider starlines that randomly connect a region of size 2^(10^20) to be implausible.
Starlines have to have an average distance of something, right? Why not 2^(10^20)?


"if there is a reason to expect human rightness to be in some sense coherent"

Alas there probably is not. Sir Isaiah Berlin speaks powerfully and beautifully of this so-called values pluralism in his book Liberty.

There are several ironies - if not outright tragedies - of life and this is one: that we don't want what we want to want, and that the things we think we ought to want often conflict with each other as well as our underlying motives. We are not in charge of ourselves and we are mysterious to our own hearts. Men and women are conflicted and, due to evolution, conflict.

I don't think I see how moral-philosophy fiction is problematic at all. When you have a beautiful moral sentiment that you need to offer to the world, of course you bind it up in a glorious work of high art, and let the work stand as your offering. That makes sense. When you have some info you want to share with the world about some dull ordinary thing that actually exists, that's when you write a journal article. When you've got something to protect, something you need to say, some set of notions that you really are entitled to, then you write a novel.

Just as it is dishonest to fail to be objective in matters of fact, so it is dishonest to feign objectivity where there simply is no fact. Why pretend to make arguments when what you really want to write is a hymn?

Wei Dai, Except that traditionally speaking an infinitelly massive universe is generally considered implausable by the greater scientific community.

But I think the greater matter is that even if it were physicially possible, it's impossible to mentally reason about as a reader of good fiction. And thus has the ability to break internal consistancy of an otherwise good store in the mind of the reader.


The story specifically asks a question that none of the commenters have addressed yet.

"So," the Lord Pilot finally said. "What kind of asset retains its value in a market with nine minutes to live?"

My answer: Music.

If your world is going to end in nine minutes, you might as well play some music while you wait for the inevitable.

Short story collections, perhaps? If you've never read, say, "The Last Question", it would be your last chance. (And if you're reading this now, and you haven't read "The Last Question" yet, then something has gone seriously wrong in your life.)

Neh. Eliezer, I'm kind of disappointed by how you write the tragic ending ("saving" humans) as if it's the happy one, and the happy ending (civilization melting pot) as if it's the tragic one. I'm not sure what to make of that.

Do you really, actually believe that, in this fictional scenario, the human race is better off sacrificing a part of itself in order to avoid blending with the super-happies?

It just blows my mind that you can write an intriguing story like this, and yet draw that kind of conclusion.

This was terribly interesting, I'll be re-reading it in a few days to see what I can pick up that I missed the first time through.

I'm not so sure we can so easily label these two endings the good and bad endings. In one, humanity (or at least what humanity evolved into) goes along with the superhappies, and in the other, they do not. Certainly, going along with the superhappies is not a good solution. We give up much of what we consider to be vital to our identity, and in return, the superhappies make their spaceships look nice. Now, the superhappies are also modifying themselves, arguably just as much as humanity is, but even if they lose (their perspective) as much as we lose (out perspective), we don't gain (our perspective) as much as we lose.

The true ending is about resisting the transformation. But they seem to accept an... unintuitive tradeoff while doing so. They trade the lives of a few billion humans for the ability to allow the superhappies to do to the babyeaters exactly what they intended to do to the humans. In fact, unless I missed something, I don't think taking this trade was ever even questioned, it just seemed that taking the trade and sacrificing the people to so that the babyeaters would be transformed exactly like humanity would have been was just common sense to them. Now, I can see how the characters in the story could perceive this choice as righteous and moral, but it seems to me to just be a another tragic ending, just in a different flavor. A tragedy due to a massive failure of humanity's morals, rather than a tragedy due to the loss of pain & suffering for humanity.

As an aside, the construction of the two (three?) alien species, with their traits, culture, and thought processes was superb.

We all have our personal approaching supernova blast fronts in T minus...

In the intervening time, everyone with a powerful enough mind, please consider engaging in scientific research that has the potential to change the human condition. Don't waste your time on the human culture. It's not worth it - yet.

Sorry I'm late... is no-one curious given the age of the universe why the 3 races are so close technologically? (Great filter? Super-advanced races have prime directive? Simulated experiment? Novas only happen if 3 connecting stars recently had their first gates be gates out? ...)

Eliezer if you're reading this, amazing story. I'm worried though about your responses to so many commenters (generally smarter & more rational than the most humans) with widely different preferences and values to what you see as right in this story and the fun sequence. I'm not saying your values are wrong, I'm saying you seem to have very optimistic models/estimates of where many human value systems go when fed lots of knowledge/rationality/good arguments. If so, I hope CEV doesn't depend on it.

If your model is causing you to be constantly surprised, then...

Well, I re-read it, and now neither ending seems so tragic anymore. I now think that there is utility in transforming the babyeaters that I didn't see before.

That said, the way they went about their supernova operation seems illogical, particularly the part about giving them 3 hours and 41 minutes. I would imagine they decided on that amount of time by estimating the chance of the superhappies showing up as more time passes times the disutility of them stopping the operation, vs the number of humans that will be killed by the supernova as more time passes, and choosing optimal time.

It seems like relatively few humans are able to escape prior to around the 8 hour mark, and, given that the superhappies gave no indications of when, if ever, they would follow (before their operation with the babyeaters was finished), the best times to blow up the star would be either immediately, if the chance of the superhappies showing up is judged to be high, or the disutility of transforming all the humans is high (relatively speaking), or they would wait 8 hours and save most of the people on the planet. Waiting about half that time seems to be accepting a significant risk that the superhappies would show up, for not much gain, while waiting about another 4 hours seems to be about the same risk again, for a much larger gain.

Still though, a very good story. I expect I'll continue to stretch my mind now and then contemplating it.

Zargon, I think the time given was how long it would take from beginning the feedback loop to the actual supernova, and they began the process the moment they arrived. If they could have destroyed the star immediately, they would have done so, but with the delay they encouraged as many people as possible to flee.

At least, that's how it sounded to me.

I know I'm way late, but I did once work out what kills you if your star goes supernova (type II anyway) while doing my dissertation on supernova physics. It's the neutrinos, as previously mentioned. They emerge from the center of the star several hours before there is any other outward sign of a problem. Any planet in a roughly 1 AU orbit will absorb enough energy from the neutrino blast to melt the entire planet into liquid rock, and this will happen pretty much instantly, everywhere. Needless to say, when the light hits, the planet absorbs photons much more efficiently, and the whole thing turns to vapor, but everyone is very very very dead long before then.

I really liked your story.
I know it's not a really insightful comment, or some philosophical masterpiece of reasoning, but it was fun and interesting.

This is, perhaps, obsolete by now.

That said, there seems to be a serious reasoning problem in assuming that this is a permanent solution. A species capable of progressing from Galileo to FTL travel in what, 30 years, seems like given another few centuries (if not much, much less) would easily be able to track down the remainder of both alien civilizations via some alternate route.

Consequently it seems like a massive sacrifice to delay the inevitable, or, at least, a sacrifice with highly uncertain probability of preventing that which it seeks to prevent. Not to mention the aliens would be rather unlikely to give humans any say in what happened after this interaction. The point about them being harder to fool is also probably true.

Perhaps I fail to understand the science, which I doubt is the issue. Perhaps flawed reasoning was intended by the author? I have to admit I was rather in agreement with the adminitrator.

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30