|The Singularity||Nature, Humanity
and AI Harmony
Ever since I heard about the Singularity, I was certain that this was the way the future would turn out. However, I recently read an article in New Scientist (no.2506 2 July 2005 pp.26-27) which presnets evidence that the idea of an exponential increase in technologuy is rather naive. This article in fact shook me up a bit, and made me think maybe I was wrong about things like an ultratech, posthuman future. So the next day (today) i went on the web and looked into these claims more closely. Here are my thoughts on this.
The New Scientist article by Robert Adler is about Jonathan Huebner (whose work is summarised in this article) argues that the rate of progress is not exponential after all, and that innovation is slowing and we will soon be in another dark age.
He forms his observations on the basis of both U.S. patents and world technological breakthroughs, per population. According to the resulting fuigures, that the rate of human technological innovation peaked in 1873 and has been declining ever since. According to his findings (and yet unpublished paper) we are heading towards a new Dark Age, an "economic limit" of very low innovation sometime around 2038.
While I can concede that maybe (who knows) the transhumanists and singularitans have it wrong, I find it very hard to envisage technological progress grinding to halt in a third of a century's time!
Also mentioned in the New Scientist article is Ted Modis, who suggests instead a gradual decline in technological novelty. This is on the basis of his mapping of the growth of complexity in the universe, a very interesting article
According to both these guys then, the transhumanists and singularitans have it wrong (i'm not saying this as a definite, just as a possibility that should be considered). However, John Smart gives a well-argued response to Huebner and in favour of the exponential paradigm (Huebner however isn't convinced - New Scientist p.27), and also replies to Modis and other critiques
The main problem with the Theodore Modis essay that I saw is that (despite its many interesting elements) the data he uses is biased towards anthropocentrism. Almost all the significant dates and events are concerned with the origin of man, discovery of fire, language, civilization, etc. And while these events indeed are important, they should not be allowed to completely overwelm the non-anthropic events and originations of complexity. Modis does not see the Cambrian explosion as a very big deal, novelty wise, yet were he to include all the extraordinary events that led to the Cambrian explosion - end of Snowball Earth, break up of Rodinia, sea level changes, rise in levels of atmospheric oxygen, change in ocean chemistry, development of "hoxbox" genes, leading to rapid evolutionary origin of all the major metazoan phyla (including Chordates/Vertebrates), development also of seaweed (as opposed to stromatolites), origin and disappearance of "vendobiota" (the strange Edicarian fauna) ...well, right there you can see eight or nine novelties, which surely puts the Cambrian explosion on the same footing as the rise of man (which likewise involved a similar number of novelties).
Another serious problem with Dr Modis' thesis is the arbitrariness of making now (or rather 1990) the turning point not just in human history, but in cosmic hsitory. I am reminded of Terence McKenna's Timewave Zero that began 60 billion years ago (that's sixty billion years ago) but only will finish in 2012 (a mere 9 years!). This is a sort of anthropocentric chauvanism in both cases, like the Medieval Christian conception of man at the center of creation. While I would certainly agree that we are currently in a very big turning point in the history of the Earth, that does not mean we are therefore at the exact mid-point of the entire cosmic history and evolution.
It is more likely to assume, as John Smart suggests and I fully agree, that if indeed there is a slowing down of novelty and technological inventiveness (and his critique of Jonathan Huebner argues that this is not necessarily the case) we may simply at a plateau, one of many, in the evolution of humanity, the Earth and the grand evolution of the cosmos, before the next big rise of complexity.
New Scientist page (nly the first few paragraphs) - also Eureka Alert page (another synpsis)
Review of "A Possible Declining Trend for Worldwide Innovation," Jonathan Huebner, Technological Forecasting & Social Change, September 2005 - critique by John Smart
Forecasting the Growth of Complexity and Change - Theodore Modis's article
Other Notable Computational Acceleration (and Selected Artificial Intelligence Critiques by John Smart
images not loading? | error messages? | broken links? | suggestions? | criticism?