Comment by jasonkolb | Hacker News Reader

Comment by jasonkolb | Hacker News Reader

Parent Full thread

Thank you, Paul, for speaking out against this cult.

It's very hard for me, as a programmer, to take this idea of a singularity seriously. I know how all of the technologies that Kurzweil is banking on work at a very low level and I call bullshit that this is ever going to happen. It would require a type of software that simply does not exist yet, something akin to a self-programming software, and we are so far away from that that it might as well be cold fusion.

I think it's much more valid to call the "singularity" the point in time when technological expansion started occurring at an exponential rate. Thus I would put the singularity at about 200 years in the past, right around the time the spinning jenny was invented, and right before the industrial revolution. Now there's a point in time that I can point to and say "something meaningful happened". This is complete pie in the sky stuff, and I'm extremely cynical that it's not just something that Kurzweil talks about to sell books and conference tickets.

Just a point that should be made in any discussion about the "Singularity": Kurzweil's isn't the only model/definition[1] of the Singularity, and some high-profile Singularitarians don't have a high opinion of Kurzweil[2] ("I've come to the conclusion that Kurzweil's worldview prohibits Kurzweil from arriving at any real understanding of the basic nature of the Singularity").

> I think it's much more valid to call the "singularity" the point in time [...]

Isn't this just redefining the word "singularity", and so making the discussion about something entirely different? So it might be better to qualify that with "industrial singularity" and the one under consideration here is the "technological singularity".

[1]: http://yudkowsky.net/singularity/schools

[2]: http://www.sl4.org/archive/0206/4015.html

I've really got nearly zero respect for Kurzweil as a theorist. He made a bunch of predictions for 2010 in his 1999 book The Age of Spiritual Machines. When 2010 came around he graded his predictions and gave himself very high marks, but when I found a copy of the book and read what he'd actually written I found that he'd had to re-write his predictions substantially in order to count them as succeeding.

As far as I can tell the only way to become popular as a futurist is to lay out predictions for the future in far more detail and with far more certainty than could every be justified.

Well we have at least a hint on how to go about it software-wise. Jürgen Schmidhuber has been working on recursively self improving universal problem solvers for decades now:

http://www.idsia.ch/~juergen/goedelmachine.html

It's... rather odd to redefine a term in use so that it's very different from what everybody else in the discussion talking about. Also, exponential economic growth has been with us since the rise of homo sapiens, but what we saw 200 years ago was a drastic shortening of the doubling period. You can also look at the invention of agriculture as another increase in the rate of doubling, and talk about how ~human level AI might cause another decrease in the doubling period. This is, at least, how Robin Hanson thinks about these issues.

http://hanson.gmu.edu/longgrow.pdf

I think we have gone through many singularity points in history. In terms of the history of the planet (or even life on the planet), the last 10,000 or so years show an incredibly fast rate of progress.

Reply on news.ycombinator.com