We need a better word than "Singularity"


Vernor Vinge (Vin-GEE)(whose 1993 novel "A Fire Upon the Deep" I published in hypertext form) coined the term "singularity" to refer to a future social and technological shift so profound and vast that those who come before it are actually incapable of understanding it.

This is an important concept, one that plays out in his novels and the writings of many others, and it needs a term. But this term has ended up not being ideal.

Scientists already have a meaning for the word of course, but it is more specific. It refers to a point where a function is undefined. For example, dividing 1/x has a singularity at 0, since 1/0 is undefined. More to the point, 1/x also increases exponentially towards infinity as you approach 0. These concepts of rapid acceleration, and the inability to extrapolate past a singularity inspired the metaphor Vinge was trying to convey.

Other forms of singularity can include any sharp corner in a function (where the derivative is undefined) and in areas within a black hole (where are normal equations of physics are undefined.) However, the non-scientific public does not understand these mathematical meanings, and thus don't quickly grasp even the metaphor.

An example of such a metaphorical singularity would be the creation of language. Pre-verbal proto-humans simply can't understand the beauty of poetry at all, no matter how much time you would have to explain it.

The "Vinge" singularity deoes not involve a discontinuity or undefined point in history. Instead, the path is continuous. You can't easily point to a specific second and say "There is the singularity where language capable of Shakespeare arose."

So the term is wrong for those who understand the mathematical meaning and meaningless to those who don't. We should seek a better term.

I welcome suggestions from readers. I think the important thing to convey is perhaps the metaphor of the "blind corner" -- a sharp, but not impossibly sharp turn which you can't see around until you get there. The ideal metaphor should also convey the acceleration of change which causes the phenomenon, and this does not. That is more akin to flying off a cliff, or the planes that turn to submarines in the new "Sky Captain" movie.


I think part of your complaint with the terms arises from your apparent view that the Singularity event will be slow enough (relatively) that people will be able to keep up with it as it occurs.

This isn't necessarily going to be the case, particularly if the Singularity ends up being driven by some form of rapidly self-improving process. In such a "hard takeoff" scenario the term Singularity comes very close to matching its mathematical roots, and indeed in such a case you will be able to look back from the future and say "yep, there was the period of 75.8 hours in 2018 when things changed".

I think also that Vinge was speaking of a specific event when he coined the term. He was speaking of the moment when a greater-than-human intelligence is created one way or another. So again we are speaking of an actual specific moment in history after which things are irrevocably different.

So in my opinion you may be grasping here for a new term that will not actually apply to what could end up occuring in the future. You're looking for a term to represent some sort of "slow Singularity" which may or may not occur, but which I think is something different than what Vinge and others may be imagining when they use the term Singularity.

I like "technological phase change".

For one thing, I don't think scientific change can run off to infinity. The speed of light and the maximum quantum information storage kind of limits the ability for information to be stored and transmitted (at rates way higher than we're currently doing, but still).

A phase change is when matter changes it's form- ice melts into water, water evaporates into steam. The rules of how ice behave, for example, follow nice, predictable patterns- right up to the point where the ice melts and turns into water. At that point new rules of how the matter behaves- rules which can not be extrapolated from the rules of how ice behaves. Note that the phase change is not, does not have to be, instanteous. Portions of the ice can be melted into water (following the new rules), while other portions remain stubbornly ice.

What we're approaching is a technological phase change- a switching over to the new rules.


Well, let's see. We could call it `the Curtain' or `the Wall'. But that doesn't convey the acceleration of change. I don't know, heck, it's so hard. How about... the Spike?

Damien Broderick

Epidemiologists use the term 'tipping point' (which is also the title of a fine book about the subject) to describe the moment when a trend goes from being unknown to being ubiquitous. Tipping points are marked by being incredibly fast acting, sometimes occuring in periods of a day or two, unexpected, and precipitated by small changes.

Its a pattern that describes everything from disease spreading to suicide rates, and in most cases, the tipping point is only obvious after it happened. After studying major disease outbreaks, epidemiologists can usually point to one event, or one small group of people that tipped a controlled infection into an epidemic. However, until the epidemic happens, one can't predict which small event is going to cause it to tip.

Societal change isn't always gradual, usually it builds slowly for a while, and then changes all at once, and it seems like the thing you are trying to name is the moment at which things begin to change all at once, the short rise time of the function. I'd go ahead and call that time the 'tipping point', when things were primed and ready (unbeknownst to most), and a small trigger event causes it all to change almost instantaneously.

True singularities probably don't really exist in a social context (and maybe not even in a physical context), since there is always a large amount of mental momentum, and time delay as communication spreads. But changes can be awfully abrupt, seeming instant when compared to years of staticity.

How about "International Talk Like a Pirate Day"? Oh, wait, that's already taken.

The point of anthesis.

If The Singularity is as close as some people say, you don't have much time left to think up and spread a new term, and the current term is pretty well established. As an Esperantist, I appreciate the difficulty of tilting at linguistic windmills. ;)

In this sentence, I disagree with the first assertion, and find the second irrelevant:
"So the term is wrong for those who understand the mathematical meaning and meaningless to those who don't."

In mathematics, "singularity" has connotations of not merely "undefined" but "goes to infinity":
with the classic singularity being division by x as x goes to zero.

That it is meaningless to people who don't understand 1/x is not very important, as 1) they probably don't care about the subject, 2) they can learn the meaning, and 3) they just need a pronouncable label that hasn't yet been taken.

I rather like the connotations of the capitalized word preceded by the definite article: not a singularity, but The Singularity.

But denying the existence of an alternate solution is easy, boring, and non-creative, so let me take a stab at it. How about The Emergence? That suggests something about the nature of what is about to happen, though by the rules we are not supposed to be able to predict it. However, perhaps there are some underlying rules that govern all these phase changes. I like Robert Wright's proposal in Nonzero:

I mentioned http://www.Nonzero.org, but let me add one more link about evolutionary game theory:

There have been a few singularities before, and may be more in the future. There's only one "next signularity" but I would not capitalize it.

A Vingean singularity is *not* "going to infinity." In fact, from the point of view of those after the singularity, they may not even be able to be sure about exactly when it was. During the great leap forward, often attributed to the rapid development of complex language, can you point to a place where things went to infinity?

The problem isn't the term or the miss use of a term rather it's a misunderstanding of an abstract. One person alluded to the fact that there was an assumption that people would be able to recognize it as it passed and yet an another tried to redefine that moment again. Where the problem truly lies is in the idea that in that defining moment there exists something that didn't exist before and the possibility that there might evolve yet something else. From an literal point of view it's impossible to view and abstract no matter what other name you assign to it. The other side of it is that same abstract can be considered and defined from a completely separate perspective for a completely different reason and indeed still be the same abstract. That is an intersection not an event horizon or singularity. Perhaps defining moment would be a better term.

Add new comment