h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

http://www.bbc.co.uk/h2g2/guide/A512902

I've put a lot of effort into this entry, and I think that I've finally accomplished an informative entry that explains the hypothetical socio-technological singularity (or "the Singularity" for short). I've fixed all errors I could find and followed all the guidelines I could find; it's 3rd-person, objective, "20th Century" instead of "20th century", et cetera.

nitpicks first...

your maths: try f(t) = 1/(1-t): f(t) is better for a function of time, and the offset by one will remove the need to think of negative numbers and suchlike.

Kurzweil and computers - is he not using the term "computer" to mean "something that computes"? Seems like he's bang on track...

To illustrate the point for non-mathematicians, some kind of table would be good. People hate equations, sadly...

Better distinction between "The Singularity" and "singularities" (in general) would be good. It confused me... On which note, mention of other singularities, such as black holes, is possibly a good thing.

Initially, I got the impression that the Singularity referred only to that of AI, specifically, rather than generally. I'm guessing that this is the "Vingean Singularity", but it's somewhat unclear from the text - at least to me.

A distinction between exponential growth and hyperbolic growth would be good: most people know vaguelly what exponential growth is, so it'd be a good starting point for an explanation, imo.

You don't really say what the figurative interpretation of the Singularity is...

How do neurohacking and nanotech give rise to hyperbolic growth? It's clear that both would be revolutionary, but hyperbolic is a big claim... Also, is nanotech only a singularity threat because of the AI aspect? Surely there's other nanotech stuff, which is just as important...

The "Singularitarians" section is likely to date pretty fast - and I'm not sure how interesting it all is... It might be better presented as a short list of external links - with a line or two on each.

--
Other stuff.

The AI Singularity being least threatening is DEFINATELY subject to debate - which I think is missing from the entry - personally, I'd say that the prospect of the human race being relegated to second place by AIs is a lot more worrying than politicians gaining another weapon of mass destruction (after all, we already have the nuke). The "active shield" concept of nanotech offers some kind of safeguards - whereas super-AIs would, by their nature, get round any safeguards we put up.

Looking at the website, the Singularity Club looks somewhat cult-ish. Worries me - hopefully unnecessarily. Reassure me...

An interesting read, and well put together, I think. As you say, it's objective and balanced - and informative too. You obviously know a lot about this topic, but I never felt I was being talked down to, which is always good. Nice stuff.

btw, does this have anything to do with the Matrix - in the scene when Morpheus is explaining the downfall of humanity, and talks about "a singular event"?

Thanks for the excellent critique, Lucinda!

I'll use f(t), but I think the offset would only make the function more complex for people who don't like equations. 1/t is probably easier to recognise as a hyperbolic function for those who know some maths.

You're right about the term "computer". I'll scrap that footnote.

I have no idea how to illustrate hyperbolic growth with a table. The point is that as t approaches 0, f(t) approaches infinity. Surely that point comes across if the reader has working knowledge of mathematics? (Correct me if I'm wrong.)

To clarify, I'll mention in some footnote somewhere that "the Singularity" refers to the technological singularity caused by positive feedback, and that "singularity" is just singularities in general. I'll also add a link to an Entry about black holes.

The Vingean Singularity is the same as the Singularity in general. That's actually an old mistake that I missed. Thanks for pointing it out.

I probably have to mention that hyperbolic growth is much "faster" than exponential growth. The reason I've tried to avoid that is that I don't want to introduce the O ("big-Oh") notation, which is the only way to compare growth that I know of. I'll have to try and just make an informal comparison.

The figurative interpretation of the Singularity is Peeters's interpretation. I'll have to clarify that.

The essence of the Singularity is the positive feedback loop created by self-enhancing AIs, computers making computers, humans hacking their own brains, et cetera. Nanotech can be used in a number of different ways: to create an intelligent computer, for neurohacking, or for whole brain emulation (or "uploading"). Thanks, I'll add that.

I wrote the "Singularitarians" section because some people wanted to know who the Singularitarians are and what they're up to. But you're right, I really have to delete the "newsworthy" details and just keep it simple.

That's a good argument for nanotech. Would you mind horribly if I quote you on that? Maybe make it a cooperative entry and write something along the lines of "One researcher thought that.." or something, like I've seen in other entries. I have no idea how to make a cooperative entry, though.

The Singularity Club IS a bit cult-ish. The good news is that it's not really a club, it's a _proposed_ society (whatever that means). I'm therefore guessing it's a one man show. I included it to show that there are some rather strange people.

Thanks again for the comments and the compliments. I'll get to work on those changes now.

Hmm. Morpheus might be talking about the Singularity, although I'll have to know exactly what he's saying there to make a guess as to whether those crazy brothers who made the movie have been reading Kurzweil's or Vinge's books on the subject.

There, I made all those changes I was talking about. Feel free to read the entry again and comment on it. I love feedback! After all, feedback causes the Singularity!

Here's how I'd do the table...

Time --- Exponential --- Hyperbolic
-2.0 --- 0.5 --- 0.5
-1.5 --- 0.7 --- 0.67
-1.0 --- 1 --- 1
-0.5 --- 1.44 --- 2
-0.01 --- 1.99 --- 100
-0.000001 --- 2 --- one million
0 --- 2 --- undefined

I think that makes it clear just how fast it grows as t->0. A graph would work too - put a comment in to the artists asking for such a beast... (I've compared 2^(t+2) with -1/t, here).

Feel free to quote me on anything... To make sure others get a credit on your entry, just put "Please credit XXXX, with researcher number 123456" at the bottom in italics - the sub-ed will see and deal with it. I've posted some thoughts under the entry - yours to use or discard.

2010, or 2100?

Thanks for the table tip, I'll add that. I'll definitely credit you if I choose to quote you (which I probably will).

How would I get the artists' attention? A comment in italics at the bottom?

It's no later than 2100 according to most people, no later than 2010 according to the Singularity Institute.

Yeah - an italiced comment will do it - the sub-ed will do the rest...

An interesting entry.

One thing - I'd change the title to 'hypothetical socio-technological singularity' as 'The Singularity' to an astronomy-geek like me conjures up thoughts of black holes.

Jimi - Good point. I'll change it to "The Socio-technological Singularity". Adding "hypothetical" makes it a bit wordy, don't you think?

or "Vinge's Singularity", or "The Vingean Singularity". Whatever you reckon...

or "Vinge's Singularity", or "The Vingean Singularity". Whatever you reckon...

Here's a question for anyone who might be reading this: Should I add the fact that a sigmoidal curve like s(x)=1/(1+e^-kx) locally resembles a hyperbolic curve? For example, f(x)=1/x resembles s(x) for large values of k and negative values of x. The difference is that a sigmoidal curve "peters out" and essentially stops growing after a while. Most conservative scientists claim that the hyperbolic tendencies of the number of people on Earth is actually sigmoidal, just like other aspects of evolution. This implies that there will be no infinite progress, but that technological progress will peter out and turn out to be a sigmoidal curve because of some theoretical or practical upper limit.

You know what? That sounds so good I'll add it anyway.

Um, I mean f(x)=-1/x, naturally. And it has to be offset a bit to really resemble s(x), but that's just a minor detail.

hmm - sounds a bit overly technical and mathy for h2g2 - bear in mind I've had complaints about the following sentence in an entry of mine:

"The entropy of some variable X, written H(X), is ...."

You could, however, add another column to your table for sigmoidal growth - to show how they are different. Then make the appropriate interpretation.

Starting to seem like you need a new subheader for "types of growth"...

Lucinda - Yes, or maybe new entries about different types of growth, curves and the "big-Oh" notation. Maybe I should just skip the sigmoidal curve, although that is an interesting aspect of the discussions regarding the Singularity. If I keep adding stuff, I'll end up writing a book. Hmmm... No, stop, this is supposed to be an entry at an introductory level, not a bible.

Lucinda, thanks again for all your help. You really like reading other people's entries, don't you? I just realised you even followed the links and read stuff in Drexler's book (the "active shield" bit) that I hadn't read yet!

Nearly missed that post about the Vingean Singularity. Much better name -- that way, people who know a bit about it will recognise it.

Ok, I've made a few changes. The entry is definitely finished now. Unless someone discovers some major flaw. Let's hope not.

Actually, I've been interested in nanotech for a while - I read about the active shield concept in "Engines of Creation" - at the foresight place. Fascinating stuff - though it might be a bit dated by now.

An entry on big-Oh notation would certainly be a nice idea - I find it a useful mode of thinking, and I'm glad I grabbed it - be good for others to know about it. Watching people trying to reason about these kind of things without big-Oh makes me concerned sometimes: I wonder what mental tools and models I might be missing out on myself...