Damien R. Sullivan wrote:
> Conclusion? Much of what we anticipate has already happened, at fast rates,
> and with the creation of sharp dichotomies. I already hear that no one person
> can fully understand a Boeing 747, or MS Excel. We can already produce
> superintelligences capable of producing things our minds aren't big enough to
> grasp. The consciousness isn't superhuman, but a human CEO makes decisions
> based on superhuman levels of prior processing, and with superhuman (in
> complexity, not just gross scale) consequences.
Doug Bailey wrote:
> Are we SIs? This question might seem silly but I'm serious. To answer this
> question we need a definition of intelligence.
We need to keep a clear distinction between Singularity::Horizon and Singularity::Transcendence. (Which of these constitutes the "real" Singularity is a matter of semantics.) Also, as many people have pointed out, the Horizon isn't at all clear-cut, so arguments about whether we're currently in the Horizon are usually "Yes" if you think of the past and "No" if you think of the future.
Questions of the Horizon become much more clear-cut if you consider the introduction of sub-Horizons which I call Speed Phases. One might view mortal history as having five major Speed Phases: Hunter-gatherer, agricultural, printing-press, industrial, and collaborative filtering. Each marked by a characteristic speed: Hunter-gatherer has doubling times measured in eons, agriculture in millennia, printing-press in centuries, industrial in a few decades, and collaborative filtering in a few years or months. We are presently in the industrial stage, of course.
Corporations are recognizably and squarely in the Horizon, even though they have some non-human capabilities. It might be argued, in fact, that the quality of corporations determine the Speed Phase. And our present-day belongs to a different Speed Phase than that of an ancient Greek, but we are still the same species.
An Internet where you can instantaneously find exactly the piece of information you'd most want to read - an Internet with collaborative filtering, Firefly tech - would make the world as incomprehensible to us, as we would be to Newton - although not as much as to Socrates. A world of neurohackers, which belongs to the same Speed Phase as collaborative filtering, would likewise be greatly altered.
Whether or not you call this a Singularity is a matter of taste. To keep things clear, I call advances in Speed Phase a Singularity::Horizon. The total remaking of the world by superintelligences is a Singularity::Transcension.
The semantic question ("Is this SI?") has now been resolved. From the perspective of an earlier Speed Phase, this is Horizon. From the perspective of a later Speed Phase, we're still plodding along. The questions of how corporations determine Speed Phases has still to be discussed, of course.
I happen to feel that "superintelligence" should only apply to Singularity::Transcendence; corporations might count as enhancements, but they aren't _super_ intelligence, which has a useful connotation of "something a LOT greater than human".
But why are corporations kept separate from Singularity::Transcendence? Instantaneous speed is probably the most fundamental characteristic of the Transcendence, from our present-day viewpoint. We've come such a long way in the past few hundred years that people who've never even heard of the Singularity find it difficult to imagine humanity in 2050, much less 2500. If intelligence becomes faster, all the change of the next ten thousand years could be compressed into hours. Even if the superintelligence is no smarter than an ordinary corporation operating at a thousand-to-one speedup, a year of technological progress can still occur every few hours. You don't have to believe in positive feedback for this to be an enormously wrenching change.
I do believe in positive feedback, and I don't think there will be anything at all left of our world - but that point is not why I'm posting this; I'm posting this to try and clear up some terminology.
-- firstname.lastname@example.org Eliezer S. Yudkowsky http://pobox.com/~sentience/AI_design.temp.html http://pobox.com/~sentience/sing_analysis.html Disclaimer: Unless otherwise specified, I'm not telling you everything I think I know.