(I prefer, of course, to assume that you’re an author, and your worlds do not necessarily reflect all of your opinions.  Even in The Cassini Division.)

Can I be the first to say thanks for joining us?

Everything you say is true, but irrelevant to the Singularity.  Sure, Siri is pretty good soft ai verbal recognition, but it’s not going to design a better version of Siri any time soon.  And nor is any other AI we have.  There’s nothing we can build at the moment that is capable of runaway self-improvement, which is the key step for the Singularity.  Current HMM and neural nets aren’t able to pass that hurdle.  They’re not just ‘a little’ short of solving the problem, they’re out by several orders of magnitude in computing power and flexibility.

Saying ‘if not for the Turing test’ is actually a pretty big exception.  It’s not necessary that the AI be able to pass Turing, or pass for human at all, but it is necessary that it be able to solve tasks as complicated as that.

If not for the Turing test, “we already do that.”  HMM, neural nets, multi-core processors, etc. Mind’s Eye is a perfect example of softa visual recognition.  Siri is a perfect example of softa verbal recognition.

“There is not the slightest indication that nuclear energy will ever be obtainable. It would mean that the atom would have to be shattered at will.” — Albert Einstein, 1932

My sense of Ken Macleod, from the books of his I’ve read, is that he has an incredibly thorough knowledge of Trotskyism in the UK and US in the 1980s, and a near-complete lack of knowledge of the radical left outside that window. He doesn’t even seem to know what happened to those same groups in the 1990s. It seems like his interest broke at his first encounter with conservative criticisms of 1980s Trotskyism, and, sticking with the assumption of Trotskyists of that era that no one else even existed on the left, he assumed that a few comments from Hayek meant the radical left was completely defeated.

As far as the Singularity and transhumanism goes, I think that in the first place, it’s a mistake to tie the two concepts too closely together. I’ve come to suspect that intelligence amplification, insofar as it may be possible, will face rapidly diminshing returns. Partly, there’s a limit to how many problems we have, that could be solved by raw calculation. And somewhere, I once saw a quip, that “Millions long for immortality who don’t know what to do with themselves on a rainy Sunday afternoon.”

Beyond that, I think one thing we’ve learned about the human psyche in recent years is how fundamental empathy is, and how we extend empathy to other living things, and even to inanimate objects. The exciting thing to me about transhumanism is that we may extend our notion of what constitutes the human community, not constrict it to nothing, as Macleod seems to fear.

We don’t know how to build one, we don’t know how to find out how to build one, and even if we could we don’t know how to make sure it does what we wanted it for.

This phenomenon (LENR) has been confirmed in hundreds of published scientific papers: http://lenr-canr.org/acrobat/RothwellJtallyofcol.pdf

“Over 2 decades with over 100 experiments worldwide indicate LENR is real, much greater than chemical…” –Dennis M. Bushnell, Chief Scientist, NASA Langley Research Center

“Total replacement of fossil fuels for everything but synthetic organic chemistry.” –Dr. Joseph M. Zawodny, NASA

By the way, here is a survey of some of the companies that are bringing LENR to commercialization: http://www.cleantechblog.com/2011/08/the-new-breed-of-energy-catalyzers-ready-for-commercialization.html

For those who still aren’t convinced, here is a paper I wrote that contains some pretty convincing evidence: http://coldfusionnow.org/the-evidence-for-lenr/

Imagine there’s no countries
It isn’t hard to do
Nothing to kill or die for
And no religion too
Imagine all the people living life in peace

You, you may say 
I’m a dreamer, but I’m not the only one
I hope some day you’ll join us
And the world will be as one

At any point along that line may exist an asymptote. Perhaps V4 makes use of all the tricks of computation and intelligence available, or V3 needs to iteratively improve itself in tiny ways for ten million years to get to the insight in V3.992 that allow V4, or we’re already at the asymptote and we can’t build an intelligence smarter than us. Perhaps the first step is millenia away. But in any case, this ‘bootstrapping’ AI concept is the core of the singularity. The ‘and then we’ll…’ is just guessing, in fact the name singularity is supposed to imply that our guesses are probably wrong.

While computers don’t last forever, on the video game example, you have Atari 2600 emulators that can run all of the games on modern PCs, and essentially the entire catalog of software there has been preserved and scattered across the internet, all 30+ year old stuff.  While working hardware is definitely still out there in droves, even if it breaks down the software is no longer tied to the hardware it was written for.

I guess that could maybe taken as a metaphor for at least part of the Transhumanist dream as I understand it, “The software is no longer tied to the hardware it was written for”.  Still pretty sci-fi, and maybe there would be some reason we discover it is impossible.  When you throw it in with faster than light travel, terraforming, space elevators, sustainable fusion reactors and Star Trek style replicators…  I’d say it seems more likely than some of those, and less likely than others?  I really don’t get the confidence many Singulatarians seem to have on the time-line they expect all of this to go down, however.  It’s so… at a speculative early kind of stage.

a better criticism would’ve been that the GNU sed most of us *nix geeks use today is (1) nowhere near thirty years old, and (2) noticeably different from the Bell Labs original. still backwards compatible with that original, true, but much expanded.

still doesn’t mean we’ll ever get to upload ourselves to the matrix, much less that it would be even remotely smart to do it were a miracle to ever make it possible (protip: don’t bet on miracles happening; mind uploading would require at least a couple). just that bit rot wouldn’t necessarily be the main reason it would be a bad idea.

but in the more general sense, the whole “singularity” business has always seemed fishy to me. having technological advance proceeding faster than the people driving it can keep up with just seems like an oxymoron, in my eyes. i’m only giving the inventors-are-all-AIs scenario a sort-of pass because i’m guessing that a human-level AI could be run on faster processors than biological neurons, outpacing us that way. but that’s really just a SWAG, too, since we have nothing like a human-level AI and do not know what their limits might actually be.

Alice in wonderland had to run as fast as she could just to stay in one place, but in mundane-land you can’t create a new advance unless you can first comprehend the problem, which implies you grasp the current state of the art. which limits the speed of progress, ultimately, to however fast the smartest humans can think. assuming we can invent something to do our thinking for us and moreover assuming that that thing will be able to think faster than we can is building our cloud castles one level too tall for my liking.

The writer had calculated that the continued existence of Afghanistan would delay the [Singularity] by six months. Millions around the world who would have had a chance of eternal bliss would be irretrievably lost to natural deaths in the interim. According to strict utilitarian reckoning, exterminating the Afghans via a nuclear carpet-bombing campaign would be the kinder course.

Singularity boffin: “My +1 Mask of the White Man’s Burden keeps falling off. After the Singularity, we will have better mask straps and no one will be able to see my colonialism”

The fact is, most computers have a useful life of 3-5 years.  The human body, 50-100.  And as for the eternity of software, when was the last time you used a ten year old program?  Or even accessed ten year old data?  And let’s don’t even get started on the possibility of malware.