From: Eliezer Yudkowsky (sentience@pobox.com)
Date: Sun Dec 01 1996 - 09:14:00 MST
> I think "subjective years" is a bad choice of terms. It requires a
> human-accessible definition. From what I can tell, in a "subjective
> year" a given computing entity can do the same amount of computation
> as a (an average? trained?) human could do in one calendar year. Pick
> a term more based in that objective measure.
Thanks! Suggestion incorporated. Simultaneously avoids a confusing
term and clarifies the model.
> >> 2. Recursive intelligence amplification.
> How can this be easy when
> >> 11. Intelligence amplification.
> these are hard? I find 11 pretty easy -- starting with chips that
> connect to your brain and help you calculate or just give you more
> memory, etc.
To explain IA you need to dip into cognitive science. To convey IA you
need more text and attention than we're likely to have plus it's
impossible anyway.
To explain that IA is recursive, you just need to point out that IAs are
better at coming up with new IAs. This only takes a paragraph.
> If you can describe three factions, make this the middle one, so it
> gets the memetic advantage neither of being first nor of being last.
> Utopia, Doomsday, and Realism might be a good breakdown.
Accepted. Realism is identical with Utopia, so ... ;) The only other
faction I can think of offhand is the "ethical to the bitter end"
faction, which says that the Powers will knock us off but that this is a
good thing. Moravec, for example.
> Nope, sorry. I'm an extropian and immortalist already, and _I'm_ not
> convinced that uploading is all that attractive -- I _like_ having
> a body. Nanotech on the other hand I like.
You'll know better as a Power, so I won't bother persuading you.
> The people you're talking to are humans. Most of them won't like the sound
> of that one bit, not to mention that it will sound millenial/apocalylptic,
> or like just another woo-woo "Strange Universe" segment.
Yes, we need to avoid this as much as possible. But the truth is that
the Universe is strange, and the materialistic Utopias of nanotech
invoke childish fantasies of omnipotence. Do you have a practical
suggestion for conveying the flavor of Beyond without running into those
memes?
> >> 10. Gray goo problem.
> > It's not all roses. The reader has a right to know, and again the
> > threat of destroying the planet can grab the attention, although
> > not quite so much after all the false alarms.
>
> I'd leave this out of an introductory letter. Alarmism is for missionaries.
Amplify, please. I'm not challenging you - I just don't understand.
How does alarmism reduce the survival potential of a meme?
> >> 13. Replacement of human society: End of History.
> > The Singularity is not one of your petty memes.
>
> Thrillsville. The end of history; just what I wanted. Foo. Show me
> how it will be the beginning of a new, better history, and I might
> want some.
Translating to memetic terms, I think you're saying that this isn't as
gripping as a bright new future that they can comprehend. I do intend
to try and - not so much make the Singularity comprehensible - convey
the essential flavor and stud it with symbolic connections to previously
present memes to make it stick. Specifically, we might present the "new
history" as a stepping-stone on the road to the Singularity; in fact,
that's what I did.
> >> 14. Singularity provides Interim Meaning of Life.
> > Ditto. One of the ways to hook a meme in the mind is by
> > making previously established memes dependent on it.
>
> According to your Web page, the interim meaning of life is to get to
> the Singularity as soon as possible. From your description, I'm in
> no hurry. Explain to people how their lives will be incredibly, even
> unimaginably enriched, or this will not sound tempting.
Good point. Add this to subject matter:
"13.5: Singularity is a good thing."
I tried to convey this with nanotech etc., but it may deserve its own
paragraph before 14-Interim.
> >> 15. Summary: History is about to end.
> > The meme as we want it to spread, once all symbolic groundwork
> > is laid.
>
> Are you aware of how easily this will get lost in the noise of crackpots
> standing on streetcorners with "The End Is Near" signs as the Millenium
> approaches?
Yes. The Millenium has been remarkably quiet so far. If we can get the
Singularity published before any crackpots get started, there should be
no interference until it's too late. Did they get 2000 and the
Greenhouse Effect confused? Still, another note for memetic
engineering:
"Immunize against contamination from 2000 meme, explicitly or
implicitly."
> I DO NOT want a majority of the human race infected with a "History
> is about to end" meme! Singularity and nanotech would be completely
> lost, and only the catch-phrase would remain, and it could be just
> the thing to cause some nut, or some group of nuts, to try to wipe
> out all humans ASAP.
So what's with all the constructive criticism? Seriously, this is
getting into a debate, but since you've challenged the project rather
than the content, that's okay. It's more a "technological progress is
about to go over the deep end" thing. Although the outline may read as
apocalyptic, what we are attempting to convey is the "WHAM WHAM WHAM
WHAM" flavor of the Singularity, in association with nanotech.
Bear in mind that memes have to spread. The "End of History" meme has
been around for awhile. Without the Singularity and nanotech memes to
support itself, it will remain silent. I don't think we need to worry
about the sting detaching itself from the bee and flying off.
> Incidentally, the Singularity page reads like a religious tract and a
> math thesis mixed freely. If that's what you want, great, but realize
> that most people won't have the capacity, much less the attention span,
> to read and understand it.
All my favorite works read like that. "Shadows of the Mind" and "Godel,
Escher, Bach" come to mind. Win or lose, you've still learned the math,
and the math (and physics) in those works tend to be of vastly more
learnable and powerful quality than those in pompous textbooks. I shall
nevertheless bear your words in mind when writing popular articles.
> I suppose I could comment further, but I've spent too much time on this
> already, and will probably just be denigrated for my trouble.
How dare you challenge my ideas! If I hadn't thought it was perfect, I
wouldn't have posted PHBBTT HAHAHAHA. Sorry, couldn't keep a straight
face.
> As a long-time Libertarian Party member, multiple-time candidate for
> office, technical writer, and budding student of social psychology,
> I've had to deal with presenting (sometimes unpopular) ideas in a
> palatable and clear way, and I draw on my experiences to try to save
> you some frustration and disappointment.
And I appreciate that. But I do wish you would be a bit more specific.
------ATTENTION READERS OF THIS THREAD------
PLEASE POST YOUR OWN SUGGESTIONS FOR AN OUTLINE.
This will allow us to mix and match the best features.
My own outline isn't powerful enough, or diverse enough,
to support a practical memetic engineering discussion.
------END OF NOTICE-------------------------
--
sentience@pobox.com Eliezer S. Yudkowsky
http://tezcat.com/~eliezer/singularity.html
http://tezcat.com/~eliezer/algernon.html
Disclaimer: Unless otherwise specified, I'm not telling you
everything I know.
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:35:52 MST