A Conversation for The Vingean Singularity [Obsolete]

A512902 - The Singularity

Post 21

iaoth

Let's not get totally off-subject with the big-Oh thing, but here's my answer anyway: I also think an entry on big-Oh would be great, but I'm probably not the right person to write it, as I don't know how to formulate it in an understandable manner...


A512902 - The Singularity

Post 22

GTBacchus

I too would like to see an entry on big-Oh notation. I've never used the stuff myself, but I hear it's great. When comparing rates of growth of two functions f(t) and g(t), I usually look at the limit of the function (f/g)(t), and maybe apply L'Hopital's rule, although with the functions used in this example [f(t) = (-1/t) and g(t) = 2^(t+1)], L'Hopital's is not necessary.

Anyway, that's all very technical and not particularly relevant to the fact that this is a very interesting and well-written entry. I expect it'll be in the Edited Guide soon!

GTB


A512902 - The Singularity

Post 23

iaoth

Thanks for the compliments, GTB. I'm pretty proud myself! smiley - smiley

Btw, I never could get a hang of L'Hopital's, plus I really hated it anyway. smiley - winkeye


"Singular event" in Matrix

Post 24

Martin Harper

I was wrong: he says "a singular consciousness", not "a singular event". Oops.


A512902 - The Singularity

Post 25

iaoth

In another thread, "Sad, mad or bad?(or Buttercup) -- U15x6/2-8+5=42!!! Word Gymnastics! A525502" (Researcher 156285) said: "VERY interesting! Although the moderator has removed two links from your singularity entry, and (if you don't mind a bit of constructive criticism) it might be a good idea to have a simple definition of singularity at the beginning, since I've only heard of it in the context of a black hole before, and I was a little confused at what you meant by it until I got to that section. I think it'd make it a little more readable. You certainly pick fascinating topics to write about! And are you a fellow maths-geek? :)" My reply: I don't mind constructive criticism at all, especially when accompanied by compliments. :) Maybe I could add a short phrase like something along the lines of "a moment of great change" wherever the reader first encounters "singularity". Or maybe an additional sentence before the whole thing. Hmm... I see that the moderator has removed a link to one company and left a link to another company untouched. Inconsistent and just strange. The other link is to the controversial Singularity Club. I wanted to show that there are weirdos that are kind of cult-ish about all this. IMO, removing that link makes the entry less balanced and objective. I wonder if I should complain. If you'd like to respond, go to the Peer Review thread at http://www.bbc.co.uk/h2g2/guide/F48874?thread=101891 That way, h2g2 is all neat and nice. :) I'm a bit of a maths-geek, but I very much prefer discrete maths. I absolutely despise the rest.


A512902 - The Singularity

Post 26

Sad, Mad or Bad? - I always wanted to be a dino, but alas, I'm just old.

Bookmarking smiley - smiley
There may be more I have to input on this, iaoth, but unfortunately I'm not feeling so hot right now, and can't concentrate. Back later smiley - smiley


A512902 - The Singularity

Post 27

iaoth

Aha, the Webmind link falls under this category: "sites which purely plug or promote commercial products or services without containing material which enhances the subject matter". In comparison, Zyvex has a lot of interesting content. Now that I think about it, the link was pretty much useless. Actually, the other link was more or less useless too, although I think it's a bit sad that readers of my entry won't be able to visit the Singularity Club and see for themselves how deranged those people seem.

Oh well. I'll remove the "URL removed" messages and just type in the link text (not the link) so that the entry is at least readable.


A512902 - The Singularity

Post 28

iaoth

Sad, mad or bad?: Thanks for the comment, I've added a more descriptive sentence at the beginning of the entry. Much better. smiley - smiley


A512902 - The Singularity

Post 29

iaoth

I made some minor changes and removed the Webmind paragraph. They filed for bankruptcy a few days ago. It wasn't a very informative paragraph anyway.

And by the way -- HELLO?!?! ARE THERE ANY SCOUTS ON THIS SITE? ANYONE???

Or is my entry not good enough? smiley - blue *sniff*

THEN SAY SO!! smiley - tongueout


A512902 - The Singularity

Post 30

Barton

I personally find the use of the term 'singularity,' dangerously confusing. This follows simply because the term singularity is used differently in, at least, two different areas of study and in each of these cases the definitions are very precise. (Math and Physics)

The result is that an inexact analogy from developmental analysis, which is being popularized, leads to its own sort of singularity of misunderstanding.

Setting all that aside, I'm afraid that I have to say that your article does not stand on its own. I can say this because, while I have a superficial understanding of hyperbolic developmental crisis through fiction and a strong understanding of positive feedback, I have not, as yet read most of the sources you cite in your article and the presentation simply does not flow without a sense that it might all make sense if I read the books you mention. But, if I had read the books you mention, I would not need your article.

Pardon me, I know this will sound harsh, but this is actually the thought that came to me at about your third or fourth footnote: This piece reads like an annotated bibliography looking for its missing article. Now that I've said it in a bad way, please let me say it better. You seem so excited and involved with what you. properly, consider a complicated and interesting topic, that you are reining yourself in too far and losing the substance for fear of losing your audience.

If you are still reading, here are a couple of more moderate suggestions. (The only response I can think of making to someone saying what I just said to you, to me about one of my projects is, "I disagree," as I turn away, muttering under my breath things that the moderator would censor me for simply thinking.)

You really must devote some time to the critics of these runaway disaster scenarios.

For instance, positive feedback is only THAT kind of a problem if it is totally unrestrained. And unrestrained feedback only exists in theory. There are resistive forces in the laws of the universe that act to restrain the increase of signal strength or transmission speed within any channel. For instance, while computer speed has been increasing at a roughly smooth pace (Sorry, I had to say it), if you examine the curve in detail, you will see plateaus not unreasonably different from the concept of quantum levels. And, if the curve isn't smooth then the analogy breaks down. Moreover, there are simply not an infinite number of quantum levels. At some point, you run out of available permissable leves and the system sub-divides.

Sub-divides? sounds like destruction to me, too. Bad analogy? I can't say.

Hold a microphone up to a loudspeaker playing back the microphone signal. What happens? Yes, but HOW loud does it get? There are severak built in limitations to the system.

Try this one, the force of gravity on earth is 32 feet per second per second. You jump out of the plane at 15 thousand feet without a parachute. How fast are you going when you hit the ground? Well, it depends on how badly you want to go fast but there finally is a thing called terminal velocity and for you without a parachute that's only about 150 mph. They system will not allow you to fall faster than that. Mind you, I still don't want to hit at that speed.

What about in space? Now we are close to a real singularity issue. You can't move faster than the speed of light in the medium through which you are traveling. In space, 186,000 mph. Which is supposed to be the absolute limit in the known universe. That's what makes a singularity. If you are falling toward a black hole, your speed stedily increases. By definition a black hole has a mass such that at some point the velocity required to escape exceeds the speed of light. Are you in the singularity? Not yet. The singularity is the boundary where the speed of your in falling due to gravity requires that you would be moving faster than the speed of light (Okay, not exactly but pretend for a minute.) You can't be moved that fast because the faster you move the greater your mass and the greater your mass the more energy that is required to move you. Do you stop? No, you accelerate slower. You can't get to the speed of light because there isn't enough energy to make you move that fast. You're still falling and for every tiny bit more you fall, the faster you must move and the more you mass. What happens? Most of the time, you start to orbit. It's theoretically possible for something to fall in in such a way that it doesn't orbit. That's where the singularity is. The point here is that the system has a real limitation. And while you can go beyond the point of no return, the curve does not go up forever.

Consider a brain uploaded into a computer. First of all we are assuming that the contents of the brain are in some real way the sum and total of the mind. We not only store memory, we simulate chemical events, and restrict the speed of the brain function so that it exactly matches the normal brain thinking speed. We supply inputs that exactly simulate all the senses of the body, all of them. We do this in such a way that the person who is being recorded has no idea that it has been done and cannot tell that he is now thinking without a real body.

All of this is done on the finest computer known to man. For the sake of argument it runs four times faster than it is possible for this person to be able to perceive a difference. Now this means that we could run the computer one quarter as fast and the man would still not know that he is now an electronic brain. So, we do that. We arrange that the computer is actually turned off for three quarters of the time. The man can't tell any difference. No conceivable function of his mind or synthetic senses can detect this condition. My question for you is, when this machine is off, is the man thinking? He would think that he is, wouldn't he?

Since all of his senses are sythetic lets record the full events for twenty four hours, but let's only turn him on one tick every year. While the machine is off, is he thinking?

If he is, where is the thought? I don't know. But, if we can slow him down then that is certainly the first step before we can speed him up. What happens when we speed him up? Either he can stand to think faster than he can sense or there is no functional difference and we are finished with human brains in computers taking over the world. If he can then we can make this man think four times faster without even breathing hard. What is he thinking about? He's a simulated man so he thinks about man things. If he doesn't then he isn't a man anymore. And we can't generalize any further because we already have machines that think faster than we do and we don't think they think, I think. They just don't think as well as we do. The whole purpose of this experience is to make a computer think as well as a man does (and hopefully it won't make the same mistakes a man makes while thinking better.)

So, we've got a fast man at best. Not a better computer. Can we make a man more like a computer? What do you want to change? What do you want to leave in? Whatever it is we are now back to AI, so what was the point of the experiment in the first place?

I have been reading Science Fiction since I could talk the librarian into leaving me alone and leaving Dr. Seuss for the other children. I've seen most of the scenarios popping up out of alarmist think tanks these days and, frankly, most of them date back to ideas that were being used in stories before 1960. That doesn't make them wrong, that just makes them equivalent to fiction to me.

All that you quoted in your article are the bare bones of theories, again that doesn't make them wrong or even uninteresting, but the old saying goes, the devil is in the details. You can't get the devil out till you consider the details. And, I guess, I could have stopped a long time ago and simply said, I want more details.

Sorry,

Barton


A512902 - The Singularity

Post 31

iaoth

First things first. It's ~called~ the "Singularity". There's nothing I can do about that...

"This piece reads like an annotated bibliography looking for its missing article."

That's actually my worst nightmare. I've read a few books that match that description exactly, and I hated them all. Ironically, they were all on the subject of artificial intelligence... Strange.

I can't say that I agree with your comments, although I can't really say that I disagree either. I refer to one book, and it's certainly not central to the concept of the Singularity. On the other hand, the gray goo problem isn't exactly so intuitive that I can just mention it and everyone magically understands. On the first hand again, I shouldn't include a full description of the gray goo problem in an entry on the Singularity.

This entry is supposed to be at an introductory level. Sure, it would be nice to have a detailed description of what Vinge thinks is the Singularity, plus a detailed description of what Kurzweil thinks is the Singularity, plus a detailed description of what Yudkowsky thinks is the Singularity, plus a... whole lot more. The problem is that it would be too long and too advanced.

My impression so far is that people who have never heard of this subject gets a rough idea of what the Singularity is by reading my entry. People who know a lot more about the Singularity than I do (eg John Smart) seem to think that my entry is informative. I guess you fall somewhere in between. You seem to feel that since you've heard about these things in sci-fi, my entry sounds like sci-fi if and only if it doesn't present more details, either pro-Singularity or pro-sigmoidal. (Or maybe Sigmoidal..? There are no Sigmoidarians, afaik... smiley - winkeye)

Yes, there's another view that says that the Singularity simply isn't possible. They agree with you and say that positive feedback only takes us so far. If you read the backlog, you'll see that I was thinking about including a description of sigmoidal curves and how they compare to hyperbolic curves. However, it seemed to me that I would only confuse the reader. On the other hand, there are some very intuitive descriptions of sigmoidal curves in your post. Would you mind horribly if I stole some of those, included them in the Conclusion section (where I keep the other Singularity criticisms) and credited you?

The point of uploading (in this context) is not simply to make humans that think faster (or think ~and~ sense things faster), although that's actually part of the goal. The main point, however, is to make transhumans (see the entry on Transhumanism, which I link to from my Singularity entry). A sentient computer program with access to its own code could, in theory, improve its own code. For example, it could get rid of a lot of evolutionary baggage; ugly "spaghetti code", (analogy alert!) hacks that Mother Nature used just to get a working prototype of intelligence. It could optimize code for speed or memory usage, maybe make some processes sub-conscious and other subject to conscious control. Et cetera!

I don't know. I've omitted details I thought were obvious, or wouldn't add much to the entry. I don't think I've missed something fundamental, and since I thought Guide Entries were ~supposed~ to be fundamental... I do agree that I could have included more pro-sigmoidal 'propaganda'. The problem was just that I didn't know how, until now.

Thus, I'm glad that you didn't just write a short post saying:

I want more details!

Barton

That wouldn't have been very constructive, so your long post is certainly nothing to be sorry about. Au contraire, it was enlightening. smiley - smiley


A512902 - The Singularity

Post 32

Barton

You are certainly welcome to use any of my ramblings that you like. I'm flattered that you found them useful.

I went back, today, and re-read your article and I find that I aven't nearly the objections I had when I first read it. In fact, I read it three times, looking for what it was that sent me hyperbolic. I couldn't find it. :/ I must have had a bad brain day.

I still feel that the article needs more detail and requires more balance than you have provided, but I *do* understand that all of this is speculation fanned into near mythology by ecstatic manics as well as crash-and-burn depressives.

I guess I *was* saying that it all sounds like more science fiction of the "if this goes on" variety and you are just reporting what's going on 'out there.' So I will refrain from doing grevious bodily harm to the messenger.

Thanks for your considerate reponse to my unwaranted ranting.

Barton




A512902 - The Singularity

Post 33

iaoth

I definitely don't feel that your post was "unwarranted ranting". (Well, most of it wasn't. smiley - winkeye)

I might be able to add some flesh to the entry, especially the "Paths to Singularity" section. For example, I've just realised that I never actually say what nanotechnology is. Silly me... smiley - silly

I'll add some sigmoidal vs hyperbolic info; probably your sound feedback example.

By the way, I'm glad you're now having a "good brain day". smiley - smiley


A512902 - The Singularity

Post 34

iaoth

Added the sound feedback example, making the entry more balanced. I'll add details later (note to self!), unfortunately I don't have time right now.


A512902 - The Singularity

Post 35

Barton

Thanks for the quote and the citation. I'm glad that you agree that more balance was needed. I'll be happy to see more detail when you can get to it.

Incidentally, you don't really need an artist. It's a simple matter to take your equations and plot them (from a spreadsheet for instance.) Then you simply save out the graph in a gif format.

The hard part is finding a way to get h2g2 to allow it to become a part of the system since we can no longer link to external graphics.

I suppose it would be possible to add a text-style graphic but it would take a lot of screen space to show any detail in how the curves diverge.

Barton


A512902 - The Singularity

Post 36

iaoth

I've made some graphs and included links to them at the bottom of the entry. Hopefully the artists can do something with them or the editor (if this entry ever gets edited... *HINT!*) could just choose one and stick it in.


A512902 - The Singularity

Post 37

Barton

I think I prefer curves2.gif. The one with the central superimposed xy axis. Of course, you /could/ add the formulas for the curves, too, but, then, I'm never satisfied. smiley - smiley

I think that adding one of these to the article would be a big help in explaining the concepts. (Yuhuu...hu?...hu?...hu Editors?...tors?...tors? Anyone there?...ere?...ere? Gee, it's real dark in here...ere...ere. And, kind of spooky...ooky...ooky.)

Barton


A512902 - The Singularity

Post 38

iaoth

smiley - laugh

I like curves2.gif too, but I don't know whether it would help to include the formulas. Doesn't hurt, I guess. Unfortunately I'm at school now, and I've got the formulas on the computer back home (an hour's drive from here). So you'll just have to wait... smiley - evilgrin There should be an smiley - evilgrin. And an . smiley - smiley Well, at least there's smiley - devil. But he looks too nice.

I've added a bit to the nanotech subsection in the Paths to the Singularity section. Now I actually say what nanotech is instead of just assuming that everyone knows. smiley - winkeye

Anything else I should add? Anyone?


A512902 - The Singularity

Post 39

iaoth

Added another curve pic -- one with formulae. However, the formula for the sigmoidal curve isn't very aesthetically pleasing. Maybe I should just have some generalised formulae? Any comments?


A512902 - The Singularity

Post 40

Barton

I don't think there's any harm in generalized formulae. The idea is to show the curves in relation to each other. But, isn't part of the trouble constructing a formula that will be in scale wit the results of the others? If these seem to messy, and I admit the sigmoid looks overly complex compaired to the others, perhaps you should pretend I didn't suggest that you paste them in there and just add another footnote with the formulae.

However, this is a symptom of the real issue and that is that the theories are modeled by simple (simplistic?) concepts and the counter-arguments state that the world/universe isn't that simple.

For instance, consider the growth of yeast in grape juice (another sigmoid process as are most growth issues. The yeast grows faster and faster till it starts to kill itself off with its own waste products. Modeling this curve requires a formula that allows for the amount of nutrient, the amount of waste, and all the other aspects of the wine making environment as well as the growth rate of the yeast.

(You've got me going on this one. smiley - smiley )

Barton


Key: Complain about this post