Frontier Channel

News and commentary about the Great Frontiers

ISS007-E-10807 (21 July 2003) --- This view of Earth's horizon as the sunsets over the Pacific Ocean was taken by an Expedition 7 crewmember onboard the International Space Station (ISS). Anvil tops of thunderclouds are also visible. Credit: Earth Science and Remote Sensing Unit, NASA Johnson Space Center

Image Credit: ISS007-E-10807 (21 July 2003) – Earth Science and Remote Sensing Unit, NASA Johnson Space Center

The Singularity Thirty Seconds From Now?

Richard Leis

,

Published.

Between 1996 and 2001, scientist of artificial intelligence and Singularity activist Eliezer S. Yudkowsky wrote and updated an essay entitled “Staring into the Singularity“. In the essay he narrows down the date of the Technological Singularity to anytime between thirty seconds from now and the year 2035. The specific date centers around the development of advanced intelligences that subsequently design follow up intelligences of a higher order, all in a very short amount of time. Humanity as we know it will end, as a new ecology of intelligences rises, intelligences so advanced that they would appear magically or godlike to us.

Of particular note in this essay is Yudkowsky’s defense of the idea that we simply cannot predict the details or the results of this future:

Ultimately, nobody knows what lies on the other side of Singularity, not even me. And yes, it takes courage to step through that door. If infants could choose whether or not to leave the womb, without knowing what lay at the end of the birth canal – without knowing if anything lay at the end of the birth canal – how many would? But beyond the birth canal is where reality is. It’s where things happen. Staying in the womb forever, even if we could, would be pointless and sterile.

Also interesting is the editing Yudkowsky found necessary since he first wrote the essay in 1996:

Similarly, computing power doubles every two years eighteen months. If we extrapolate forty thirty fifteen years ahead we find computers with as much raw power (10^17 ops/sec) as some people think humans have, arriving in 2035 2025 2015. [The previous sentence was written in 1996, revised later that year, and then revised again in 2000; hence the peculiar numbers.]

Valid predictions based on extrapolating today’s trends have begun to fail not because they do not come true but because they happen sooner than the date expected. Progress of particular technologies tend to follow a curve that grows exponentially at first but then plateaus as the technology hits physical limits. However, new technology comes along that leapfrogs over the old technology and starts a brand new curve, often with an even faster growth and steep slope. The power of convergence and new paradigms are often ignored by those who see no such progress as work.

%d bloggers like this: