- The end of Moore's Law and failure-to-materialize of the Technological Singularity*
*This is more of a refutation of another popular megatrend idea than a trend in itself
Yes. I am in real agreement here. I'll assume you read some of Kurzweil's tome? I found it kind of easy to find problems with the way he defines even simple concepts like computational power.
Kurzweil is an interesting guy. If you can find his earliest book,
The Age of Intelligent Machines, you can see a lot of things that he predicted correctly when they looked like Star Trek fantasy to most people.
I don't have that. I do have "Spirtual Machines" on my bookshelf but it's currently unread. I've been at a few lectures by Kurzweil (that's where I got my copy of "Singularity"). I had wanted to ask him about quantum level effects in thought (not necessarily Penroses ideas) but they cut off questions right before I got to the mike.
You can also see the first of his repeated predictive mistakes there (in that book, like other writings since, he incorrectly predicted that voice recognition was just a few years away from dominating computer user interfaces).
That's interesting timing. I recall many discussions with people over the Newton platfom that Voice Recognition would eliminate the need for a stylus on smaller devices. Interesting how UI went sideways from there. Keyboards are the predominant way to work with a mobile platform.
Most of the Singularity ideas Kurzweil popularized were first discussed on the Extropians mailing list starting in 1991. And -- interesting Bitcoin connection -- the early Extropians were also connected to the Cypherpunks movement, which started discussing secure digital communications, decentralized payments and contracts, and the other electronic building blocks of radical libertarian utopia, nearly 20 years ago. Vernor Vinge, who invented the Singularity concept, also practically invented the Cypherpunk vision in his 1981 story True Names.
Yeah I've read some of Vinge and I know he originated the term. Some of his other side ideas seem plausible too. The idea of "code archeology" mining for code that does one thing or another. Given the way that Open Source allows the free (in variously defined senses of the term) licensing of code and the internet allowing easy access. Cut and paste coding is very much a defacto practice (so much so it has a some proprietary vendors worried)
As an aside. Do you read Gene Wolfe - IIRC he mentions Vinge in-story in Fifth Head of Cerebus.
I think Kurzweil is wrong about the evolution of coming technologies, particularly about how fast they're going to evolve. I think that most of his predicted technologies could some day come to pass, but I very much doubt that they're all going to arrive in the next 50 years like he does. In a way I almost feel sorry for him, because I think many of his predictions could come true on the given time scales if everyone were as bright and educated as he is. He is an exceptional person who has made the unexceptional mistake of assuming that other people are basically like him.
I agree the fifty year date is unrealistic I also think in some cases he confuses some concepts. i.e. Transistor Count is not the same as task performance. So I'm not so sure that meeting his goals just boils down to just having more of him though. I think the saddest thing for him anyway, about his predictions is that he pretty much excludes himself from being one of the preserved. These particular transhumanist ideas tent to remind me of the cryonics "boom" (if you can even call it that) in the early 90s and perhaps just as wishful.
The fastest way to gage Kurzweil's predictive ability is to look at his predictions from The Age of Spiritual Machines. He made a lot of predictions about future decades, and we are now living in the earliest of those predicted eras. My evaluation is that he was right about some things, wrong about more things, and you can see a pattern to his mistakes. He ignores social, political, and even scientific factors that could derail his predictions, like wars, fashions, legal liability, investor expectations, the complexity of neurobiology, and so on. Even though he is wrong in the particulars I think it's good to have at least one prominent figure who is really excited about the future and believes in the power of technology to improve everything, lest we all get sucked into a gloomy vision of the future where the only excitement comes from calamities.
To a point I agree but personally I'd rather read about more near term stuff. For me anyway learning about even something like islet cell therapy I think is both exciting and keeps me grounded in the idea that "science is hard" and often slow.