Talk:Technological singularity/Archive 2

From Wikipedia, the free encyclopedia

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Introduction rewrite

(This is in response to Piotrus's comments above.)

"[...] both of make the definition shorter, and I could live with them, as it wise to make the definition as KISS as possible but limiting it just to societal change is I think a bit to specific, I would suggest droping the adjective and just go with 'change' or 'change in our enviroment.'"

Yes, I mostly just wanted to make it shorter. I do like change in our environment better, or maybe even global change or something equally all-encompassing.

"[...] you limit to 'advent of AIs', and from what I read, not everybody agrees AIs will be invented first - some speculate that nanotechnology will be required to reach that level."

I intentionally didn't specify AI, just superhuman intelligence, whether this is produced by creating strong AI or augmenting existing human intelligences (through nootropic drugs or cybernetics or whatever). This is the key to the Singularity: that there is something smart-than-human driving progress, and it can make even smarter entitites, and so on and so on until things go kaplooie. Vernor Vinge wrote in his original Singularity essay, "Von Neumann even uses the term singularity, though it appears he is thinking of normal progress, not the creation of superhuman intellect. For me, the superhumanity is the essence of the Singularity. Without that we would get a glut of technical riches, never properly absorbed." This essay also begins with these lines:

"The acceleration of technological progress has been the central feature of this century. I argue in this paper that we are on the edge of change comparable to the rise of human life on Earth. The precise cause of this change [emphasis mine] is the imminent creation by technology of entities with greater than human intelligence." [1]

Also, the first sentence of What is the Singularity? from the Singularity Institute: "The Singularity is the technological creation of smarter-than-human intelligence." (From What is the Singularity?)

I do understand that this aspect of the Singularity is left out in some writings, perhaps most importantly Ray Kurzweil's, who sees the Singularity as rapid technological change following a roughly exponential trend that's been going on for ages and not needing any special cause to kickstart the process. But Vinge is more prominently associated with the Singularity than Kurzweil, and the article should use the meaning he uses in The Coming Technological Singularity, while acknowledging that others have taken a different spin on his ideas (hence the last paragraph of the intro). I'd like the article to differentiate more clearly between the Vingean and the Kurzweilian models, as in my own experience this is a major source of confusion surrounding the Singularity.

"Finally, you replaced the 'pre-Singularity humans' with 'present day humans' - this is an error, as many theories argue [...]"

No objection. I think either one fits and am not quite clear what you're getting at with the accompanying argument, but it seems like this is just semantics. Pre-Singularity is fine.

Sorry if I hadn't made myself clear. From what I understand, it looks like this: modern/presend day humans will not be able to understand Singularity and post-S world (or at least, major parts of it). However, those very humans (i.e. alive today, like you and me) can be 'upgraded' during/after S, so they could keep up with the pace of changes. Thus there is the difference between 'present-day humans' and 'pre-Singularity humans', as 'present-day humans' in Singularity era can actually be 'post-singularity humans' (or post-humans beings - phoebe - as my favourite writer, Jacek Dukaj calls them :). Yes, the difference is small, but I want to avoid any confusion in the definition if possible. --Piotr Konieczny aka Prokonsul Piotrus ^Talk 23:05, 20 July 2005 (UTC)

"One last thing: why do you think that 'evolution of society' should be a subsection of the Kurzweils' law section? I see little connection, at best I'd do it the other way around."

Having so thorough a rehash of the history of human society and theories of people that don't even speculate as far as the Singularity seems not worth including at all to me, but if this history ties in at all its through a Kurzweilian model of the Singularity showing how the time between various technological advances diminishes and culminates in singularity as the oh-so-contentious graphs discussed above depict. This current arrangement isn't ideal, I agree. I was half motivated by the Kurzweil graphs that tie in best with the "Evolution of society" section and wanted to maintain NPOV by having them under the Kurzweil section so they seem more like depictions of Kurzweil's beliefs, diminishing the bias much of the article has in favor of his theories. --Schaefer 21:13, 20 July 2005 (UTC)

HTML Code

I noticed one of the latest revisions to this article used some html code (div tags) in order to "reduce extraneous whitespace". The result is that the "Prominent voices" and "See also" sections are side by side. Is this usual on wikipedia? I don't think I've ever seen an article using this technique... If it's not a standard thing, I'd say we shouldn't do this. --Rbarreira 02:00, 21 July 2005 (UTC)

Never seen it before, but it looks interesting. But I am afraid it may cause problems on some computers/browsers, I think that html tags should be avoided whenever possible. --Piotr Konieczny aka Prokonsul Piotrus ^Talk 10:09, 21 July 2005 (UTC)

"But I'm afraid it might cause problems on some computers/browsers." In the unlikely event someone reads this article with a browser that doesn't support CSS, it just shows up as at did before. It doesn't interfere with text-only browsers or screenreaders for the blind. That's the beauty of CSS over HTML markup. As for this being the first, it isn't. I've seen it other places, but this version of the article Philip Glass is all I can find right now (it's since been taken out because both of the sections have been expanded out so far and given suheadings). I know of no WP policy against this, and have made some effort to search for one to make sure. If anyone else has better luck in finding one let me know. --Schaefer 16:49, 21 July 2005 (UTC)

Maybe it's not forbidden, but IMHO it's unnecessary and a bit weird for the people who are already used to the usual wikipedia appearance. But if I'm the only one who thinks this, we should of course let it be like that. --Rbarreira 21:57, 21 July 2005 (UTC)

I agree with Rbarreira. Function before form. If we get too far off the beaten trail, we're just going to confuse people more. Further, I see absolutely nothing wrong with having one section after another: indeed, for layout and ease of reading, whitespace is your friend. (I also agree, however, that HTML can be fine, but only if it serves a function, and not just a layout form that achieves no further enlightenment for the reader.) I'm sure many people would agree that the HTML tricks is more appealing, but it causes confusion too, such as "are these two concepts somehow connected?" --MShonle 22:56, 21 July 2005 (UTC)

I just want to make sure that Schaefer doesn't get me wrong... He's one of the best contributors to this article and I appreciate his work a lot, but I just found this change a little odd. Anyway, I think more people should say what they think about this :) --Rbarreira 23:45, 21 July 2005 (UTC)

Unsourced

Schaefer, I think if we wanted to comment out all unsourced info, we would have no material at all. Besides, I think it is better to move such text to talk then to leave it invisible in the main body. --Piotr Konieczny aka Prokonsul Piotrus ^Talk 10:18, 21 July 2005 (UTC)

Here're the bits I've removed for being unattributed:

"Another criticism is that "singularity" events are a constantly recurring event. Any historian can cite a time in history where the people would be unable to imagine the world we live in today. In a Harper's Bazaar article from 1854, the world of tomorrow saw people being shot short distances in oversized cannonballs (pre-plane) while traversing the country in underground trains (pre-car). Today is a singularity to anyone living in 1854." Reasonable argument, but is likely the personal view of the contributor rather than a notable critic. Needs to be rewritten with weasel terms if it's to go back in. -- Schaefer 17:12, 21 July 2005 (UTC)

You mean 'without' I think? :) --Piotr Konieczny aka Prokonsul Piotrus ^Talk 18:53, 21 July 2005 (UTC)

"This linkage to the modern Dispensationalist Christian Rapture belief seems to be supported by an existing correlation between the atypical popularity of Dispensationalism and Technological Determinism within the United States, as compared with many other developed nations. It may be that current popular versions of the Technological Singularity may be nothing more than culturally specific remixes of the standard, enduring eschaton myth dressed in science fiction tropes." Should be attributed to Ken MacLoed if he in fact said this. It appeared with a quote from one of the novels, so it implied he did without really saying it. --Schaefer 17:12, 21 July 2005 (UTC)

Schaefer has done serious violence to the article by hiding his POV behind ridiculous standards. I would support any one who would revert all of his changes. In the meantime, I'll just put back the content I've provided since so many other edits have occured after Schaefer's changes. Perhaps I would assume better faith had Schaefer used the talk page before damaging so many people's work. --MShonle

My personal philosophy is that if the talk page doesn't already have a discussion about a section of text, then an editor can do as she pleases and if anyone objects, the changes are reverted with no harm done, and a talk page discussion starts so compromise can be reached. I don't see "violence" done in taking questionable text out of the article for a few days at most (or a few hours in this case) while we discuss its neutrality. But in the interest of maintaining the peace, I'll try to be less bold in the future. --Schaefer 01:18, 22 July 2005 (UTC)

For the record I didn't mean "all of his changes to the article"; I just meant the single change that happened to delete a lot of text that was mostly workable instead of requiring abandonment. --MShonle 20:15, 21 July 2005 (UTC)

Heh, thanks for that clarification. Scared me for a second there. --Schaefer 01:18, 22 July 2005 (UTC)

Portrait photos

I don't see what having photos of Vinge, Kurzweil, Bell, and Toffler adds to the article. Their physical appearance isn't relevant to the topic, and this article already has an abundance of images. I would like to remove them if none is strongly opposed. --Schaefer 22:09, 25 July 2005 (UTC)

I'm not strongly opposed to their removal, I just think it's nice to put human faces to the concepts presented. My poor brain processes information much better when presented by "people". --Rose Lacy 00:26, 26 July 2005 (UTC)

It's common to add portaits to rekevant articles where space permits. We have space- there is still no image overload (IMHO) and it's nice to see people who invented most of the concepts we are talking about. I'd like for them to stay. --Piotr Konieczny aka Prokonsul Piotrus ^Talk 09:46, 26 July 2005 (UTC)

My opinion is that Daniel Bell and Alvin Toffler are not relevant enough for this article to deserve pictures. They're only mentioned once (exactly in the place where the pictures are). If they deserve pictures, I think many more people who are mentioned in the article would too. Since the place where their pictures are is already the most pic overloaded one, I would suggest removing those two pictures. What do you people think? --Rbarreira 13:47, 26 July 2005 (UTC)

Agreed. I'm taking out those two for now, but leaving Kurzweil and Vinge in. --Schaefer 00:14, 27 July 2005 (UTC)

Logarithmic graphs

Some comments from the perspective of the Logarithmic timeline article(s):

1. The graphs here have excessive unnecessary borders, which should be trimmed.
2. You can't predict anything from a log timeline.

I did an extensive one and found that the Big Bang, classical Greece-Rome, and the countercultural Sixties all offered 'inflection points' where change was accelerated compared to before and after. If a tech singularity is coming (which I accept, and have spent my life working for), what follows it will distort the logarithmic progression. --robotwisdom 11:34, 28 July 2005 (UTC) I believe that with the growing visual literacy on the Internet, inclusion of imagery is not just preferable, but essential. We are moving into a richer media online and wikipedia needs to embrace this as a challenge to the hegemony of text. Phil McRae

Just to play with this concept I created a log-log graph of major extinction events.

The reason I chose extinction events is that they are the least anthropomorpic set of sequences I could find. Classifications of various families and species tend to have a very human centered POV, with a lot of focus on those families which give direct decent to man (mammals, apes etc) rather than the vast number of other species. From a sharks POV all these mammals are just another type of tasty snack!

So what do you think? Does this graph fit with that nice straight line hypothesis? Note x-axis is reversed for technical reasons. --Salix alba 16:11, 16 January 2006 (UTC)

For what its worth heres the raw data, derived from Timeline of evolution

        years ago       time to next event 
Life    4.00E+09        3520000000
1st     4.80E+08        40000000
2nd     4.40E+08        90000000
3rd     3.50E+08        100000000
4th     2.50E+08        100000000
5th     1.50E+08        85000000
6th     6.50E+07        64999999
Now     1.00E+00

--Salix alba 16:19, 16 January 2006 (UTC)

Kurzweil vs Technology Review

Mass use of inventions: Years until use by a quarter of US population. Some believe growth curves like these may soon flatten or become more linear.

I found the TR data suspicious (radio (9) and television (6!) - 6 years after the first consumer tv sets were released, half of the US population owned them??? doesn't it raise your eybrows?). I then attempted to verify them (and Kurzeweil data as well, to be fair). I checked one parameter - television - and found that Kurzweil is correct, TR is not. Thus I'd like to ask if anybody would object if we remove the TR table (btw, anybody has a direct link to that article?).

My findings: the problem with this is agreeing on the date we start counting from. Both articles are correct when they quote the information about penetration in given year, however TR sets a very arbitrary date for 'start' of the penetration (or rather, I have no idea why they chose the year they did). The data for penetration itself should is fairly easy to get: the simplest and best solution would be to go to statistical yearbooks which should contain information on number of households owning radio/tv (they may be online at [www.census.gov www.census.gov]. I didn't go there but googled them out at:

• Number of US households with TV - crossed 25% between 51/52 and 50% between 53/54. Which means that the start date (0 penetration) of Technology Review table is 1949. Which means you can take the TR table and throw it in the bin, cause it is obviously a very arbitary date and the table is misleading and cannot be used for comparison with Kurzweil graph, because:
• The first tv broadcat goes back to 1928 (from our article on television): "The first regularly scheduled television service in the United States began on 2 July 1928". We can assume that it means that some (prohibitively expensive) tv sets were available in stores (or on order). Our television article states that the first commercial licence for tv station was granted in 1941 ("the FCC saw television ready for commercial licensing, with the first such licenses issued to NBC and CBS owned stations in New York on 1 July 1941"). I guess it is safe to assume that tv sets were much more widely commercialy available at that time. If we take year 1928, we have the 50% penetration number of about 26. If we go with 41, it's about 13. The number 6 is bogus, I can't find a single important event that would indicate it should be adopted (feel free to correct me, of course).
• Kurzweil's graph states that the TV was invented in the second half of 1920s (consistent with first broadcast date) and it was used by a quater of US population (don't know if he equated this with households!) in about 25 years (remember his graph is a reverted logarithmic scale). If he meant households, his graph is correct (in that part).

Some other interesting facts I googled out:

• This article (pdf) states: "In 1950, slightly fewer than 10 percent of U.S. households had a TV, but scarcely five years later that number jumped to 65 percent. " I assume it made no distinction between color and black and white. So it means that it took 20+ years we had less then 10% market penetration, then in five we have a sudden spike to 65%! A trend confirmed from the table I linked above, and fairly expotential-looking to me :)
• Some date on color tv: [2] and [3] indicate that the mass production of color tv sets begun around 1954, although the invention was made before 1940s (see our article again). Our article on television also states that: "as late as 1964 only 3.1 percent of television households in the U.S. had a color set". So, as far as color tv goes, the penetration of 3% (!) took at least 10+ years (assuming we startf from mass production date) or over 20 (if we start from invention date). Just another nail in the coffin of the '6 years' date, and I wanted to make sure they didn't mean color tv.
• Some nice statistics and links if you want more info: [4], [5]
• One more illustriation how the table is misleading: it lists DVD players before Internet, even through they are a much later invention (1996 vs late 1970s [I don't want to go into details, see History of the Internet for them]). Btw, note that Kurzweil graph does not refer to the Internet but to World Wide Web, which he correctly puts at begining of 1990s.
• And one more. It can be argued that wireless phone is older then the Internet, with first models going back to 1940s/50s. Although it would be going for the trivia, and usually it is assumed that mobile phones go back to early 1980s. Definetly it is again older then DVD player, making the number of the table misleading (becouse it is not a time series and so it cannot be used to compare to a timeseries and conclude there is no trend!).

If I had more time (which I don't) I guess I could do a similar analysis for other inventions (it was kind of fun). Still, I'd motion for removal of the Technology Review table, it seems erroneous in at least one part. --Piotr Konieczny aka Prokonsul Piotrus ^Talk 17:12, 25 September 2005 (UTC)

Article is way too much about Kurzweil

The article is way too much about Kurzweil, to a point of craziness. The Kurzweil fans that did this to the article should create a separate article about Kurzweil's singularity ramblings, and now would be a good time to do it since his book just came out. --Tempshill 17:48, 29 September 2005 (UTC)

Feel free to add or move content. Deleting it because it doesn't suit your POV borders on vandalism, especially as the graph issue was discussed at some lenght here and the community consensus was to keep those that now remain. --Piotr Konieczny aka Prokonsul Piotrus ^Talk 21:52, 29 September 2005 (UTC)

I don't know about creating a new article for Kurzweil's Singularity opinions, but I agree that the article as it stands overrepresents him. I'd be strongly in favor of moving some of the content to article fr Kurzweil himself, or possibly the articles for his books. --Schaefer 04:46, 30 September 2005 (UTC)