Transumanar

Cranks and (deleted)

Dale Carrico, the chivalrous champion of lost causes who likes to criticize others but does not like to be criticized by others, has just informed us that all transhumanists are cranks.

A transhumanist, you should know, is not the Wright Brothers or Thomas Edison, but “the dot-eyed crank in the basement” who believes in “genetic and prosthetic medical techniques or brain scanning and modeling techniques that will transform some of us into imperishable robots and end human mortality” (now, this is really disgusting!). Our “hopes are essentially faithful and not scientific and the hysteria and false certainties Superlativity mistakes and peddles as hope are worse, essentially fundamentalist”. And, of course, he does not need to discuss “with those incapable of grasping that “uploading” is Robot Cult nonsense”.

Well, I have always appreciated Dale for saying what he thinks. Too bad that he does not bother to offer us a scientific explanation of exactly why “uploading is Robot Cult nonsense”. It is nonsense, of course, because Dale says so. The brain, of course, is not a computer but a gland, and the mystical and ethereal mind essence that it produces will never be amenable to corporate-militarist (???) engineering analysis.

By the way, a very interesting article appeared last week on Technology Review (the TR of MIT, not the TR of the Robot God Cult), titled ”A Working Brain Model - A computer simulation could eventually allow neuroscience to be carried out in silico”, has some interesting news: “An ambitious project to create an accurate computer model of the brain has reached an impressive milestone. Scientists in Switzerland working with IBM researchers have shown that their computer simulation of the neocortical column, arguably the most complex part of a mammal’s brain, appears to behave like its biological counterpart. By demonstrating that their simulation is realistic, the researchers say, these results suggest that an entire mammal brain could be completely modeled within three years, and a human brain within the next decade”.

Posted by on 12/03 at 05:24 PM

Dale Carrico, the chivalrous champion of lost causes

I’d enjoy hearing about the “lost causes” you think I advocate. It will be intriguing to compare them to the causes we claim we share politically when you are disputing my claims about right-wing bias in superlativity.

who likes to criticize others but does not like to be criticized by others,

Which is why I continue to engage with you here, since you’re so non-critical of my stance, I suppose.

has just informed us that all transhumanists are cranks.

It’s you all who are demonstrating it, I simply provide evidence of the self-immolating spectacle.

A transhumanist… believes… “techniques that will transform some of us into imperishable robots and end human mortality” (now, this is really disgusting!).

I don’t claim this is disgusting, I claim that it’s dumb.

of course, he does not need to discuss “with those incapable of grasping that “uploading” is Robot Cult nonsense”.

And yet here I am.

By the way, a very interesting article appeared last week...

Science is science, subsequent superlative spin and attributions are another matter. Let me know when you’ve “uploaded” into cyberspace or are wallowing around in utility fog hog heaven… I’ll be happy as always to exchange pleasantries. But I won’t be holding my breath, though!

Posted by Dale Carrico on 12/03 at 09:07 PM
Again, you seem like you aren’t arguing against what he is saying. The difference between transhumanists and cranks is that perpetual motion machines have been proven impossible, while mind uploading has not. Mind uploading could be hype or fringe science, but it is not outright impossible. It could also be pseudoscience, but I don’t see much in common between Robot Cultists and people who claim that Laetrile can cure cancer.

Posted by remterbacysi'ox on 12/03 at 11:43 PM
Oh, and about the “lost causes"--a blogger who I mostly agree with (Mencius Moldbug of <a href=http://unqualified-reservations.blogspot.com>Unqualified Reservations</a>) thinks that Republicans are the winning party because they have more money. They aren’t winning now, but they win in the long run because money gives them more resources to get votes.

The question isn’t whether world federalism and technoprogressivism is good or bad, but whether it is politically possible.

Posted by remterbacysi'ox on 12/03 at 11:47 PM
‘Dale Carrico, the chivalrous champion of lost causes’.

Is that really fair? You think uploading is feasible, Dale does not. In all honesty, I think we still know so little about how the brain performs its magnificent capabilities that any opinion is as valid as any other at this point in time. I happen to know of many serious and rational scientists who think uploading may be feasible. I also have no end of books, written by very respected neuroscientists, who share Dale’s view that uploading is plain nonesense. Who is wrong? I don’t know. But I DO know that people have a tendency to fall for ‘confirmational bias’, whereby you tend to seek out evidence that conforms to your prejudices, rather than being an unbiased investigator of the facts. Both Dale and Giulio should perhaps ask themselves how much confirmational bias is clouding their judgement?

“these results suggest that an entire mammal
brain could be completely modeled within three years, and a human brain within the next decade”.

In a previous post, I referred to the ability to obtain second opinions on every stated claim. In the case of the Technology Review story quoted by Giulio, it so happens that the story itself features just such an opinion, from Kristof Koch, professor of biology and engineering, Caltech.

‘Any claims that the human brain can be modelled in 10 years are so ridiculous they are not worth discussing’.

Yeah, well, we all know that the human genome project was thought to take 1,000 years after a decade of work had completed 2%, but then subsequent generations of scanning technology, following the Kurzweil Curve, allowed the remaining 98% to be sequenced in 5 years. Who will have egg on their face in this ‘the brain will/will definitely not be modelled in a decade’?

Time will be the ultimate judge.

Posted by on 12/04 at 12:01 AM
Republicans are the winning party because they have more money.

Progressive taxation + public welfare programs that benefit majorities = Democrats are the winning party.

Posted by Dale Carrico on 12/04 at 12:51 AM
Republicans can block this. Even if the Democrats can get people angry at the Republicans, the Republicans can still maintain a large group of people who don’t want progressive taxation (as they are doing right now) and use them to block any legislation that comes up. This helps corporations, and they will give some money back.

Also, I suggest that you read some of that blog (especially the “How Dawkins got pwned” series). His views seem to be diametrically opposite from yours.

Posted by remterbacysi'ox on 12/04 at 01:56 AM
Newsflash! Everybody hates the Republicans, they are likely to lose the Congress, the White House, the governorships, with nothing but the intransigent white racists of the South in their corner.

Leave it to the futurological brain trust to be the last people on earth to continue to take movement conservatism and market fundamentalists the least bit seriously.

p2p democratization in the form of the Netroots drove a change election in 06 (largely betrayed by inside the beltway conventional wisdom, but giving way in the face of people powered donor support and rapid media pushback) and will drive primary challenges against corporatist incumbents in the DNC. Progressive dems beholden to an empowered Netroots constituency will push through universal health care, campaign finance, net neutrality and media disaggregation, and both medical and green r&d;. (I call this the emerging technoprogressive mainstream.)

Republicans capitalized on white racism in the aftermath of civil rights to render dems vulnerable but demographics have undermined that strategy fatally. In the aftermath of the New Deal big business organized (see Mt. Pelerin, see the American Enterprise Institute, see the CHicago School, see Grover Norquist) as did religious conservatives that had not hitherto had much of an organized political existence: that alliance is falling apart as social conservatives recoil from corporatist secularism and corporatists disdain culture wars that inhibit lifestyle diversity that provides profit opportunities.

This is a quick and dirty and very US-centric sketch of complexities that are much richer, and much more encouraging for dem-left progressives the closer you look and the more planetary your perspective, in my view.

It isn’t that hard to explain to people that taxes are the price we pay for a government that works and that people who hate government can’t be trusted to govern. Majorities supported traditionally left positions on issues even at the nadir of dem powerlessness. White racism and commie paranoia really can only work for so long, even with bought and paid for media hacks exclaiming that these foolish slogans are hot ideas for three decades straight.

My point isn’t to paint a rosy picture or offer guarantees, but the fact is Republicans are killer clowns and scoundrels and bullies who destroy everything they touch and everybody hates them and there is every reason to think they can be defeated by a diverse dedicated citizenry in even notionally democratic polities.

As for Dawkins, I really like some of what he says, but I rather dislike quite a lot too. I’m a cheerful atheist and technoscientific enthusiast of many years standing, but I have no patience for atheist militancy or scientistic reductionism. I can also walk and chew gum at the same time.

PS: Robot Cultism is still dumb.

Posted by Dale Carrico on 12/04 at 03:38 AM
Extropia: “‘Dale Carrico, the chivalrous champion of lost causes’. Is that really fair?”

It is most certainly not fair, and it is not what I think of Dale anyway (actually I like him). But we are playing a nice verbal game here, where he feels free to attribute to me positions that I never held, to conclude that transhumanists are crankish Robot God cultists. So, I feel free to do the same.

“Both Dale and Giulio should perhaps ask themselves how much confirmational bias is clouding their judgement?”

This is always a healthy attitude, and actually I do read the arguments of the neuroscientists who share Dale’s view that uploading is plain nonsense. I just don’t find them correct.

I think uploading is possible, in principle, almost by definition. Here I am following Sherlock Holmes’ “When you have eliminated the impossible, whatever remains, however improbable, must be the truth”. What is impossible, for me, is the supernatural (it is a logical impossibility like a triangle with four sides), so I assume that everything under and beyond the stars is a physical object that must obey the laws of physics, however weird they may prove to be, and can in principle be reverse-engineered and improved upon once we have mastered the engineering applications of these laws.

Note that, if Dale said something like “immortality (read indefinite lifespan) would be a beautiful thing, but sadly I do not think it is a practical possibility”, I would not engage in any argument with him but just conclude, as you do, that “Time will be the ultimate judge”. But Dale actually _celebrates_ our mortality and vulnerability.

Posted by on 12/04 at 07:20 AM
Concerning Democrats and Republicans, I hope and also think that the Democrats will win the next elections, and that this will have positive consequences for the US and the rest of the world. But I think Republicans will go back to power after the usual 8 years and so on. It is just too easy to win elections when you can blame the party in power for everything, especially in a two party system.

I do not share Dale’s views of Republicans in general, but I certainly share his views of the current Republican administration.

Posted by on 12/04 at 07:29 AM
Extropia: in the Technology Review article, Christof Koch says: “any claims that the human brain can be modeled within 10 years are so “ridiculous” that they are not worth discussing”.

He is talking of engineering development timescales, not (necessarily) of theoretical feasibility in principle. Actually, I would be _very_ surprised if the human brain were modeled within only 10 years.

But as you correctly say, engineering development timescales turned out to be much shorter than expected for human genome mapping. This example should show that “ridiculous” is a very dangerous word to use, because it can bounce back and make _very_ ridiculous those who used it.

Posted by on 12/04 at 07:58 AM
from the earlier comments:
I don’t find this so shocking. A simple analogy is: if you backup a program’s source code and its database, recompile and install it on a different computational system (different CPU, different memory hardware, wiring etc.), for which a compiler for the same source is available, everything should work exactly as before, so for all practical purpose this specific construal of information is indifferent to the material mode of its instantiation.

This analogy, which should of course be developed in more detail (actually it has been developed in much more detail by many authors), will probably seem “reductionist” to you because it does not take into account some nebulous property of the brain-mind system, but I think it is basically correct.

High-level languages were designed for just that, along with human-readability. Quite a bit of efficiency and simplicity was sacrificed in the process. It’s not true for native machine code (or assembly languages where just human readability was the goal), not unless you are ready to emulate the whole “source” machine, (and its firmware and OS).

Now, are there a slightest indications of evolutionaly pressure for “mind as a C++ program” versus “mind as an assembly code for INTERCAL chip”? Hovewer indications of contrary pressure at work exist. Cells don’t have anything like “program in C++”, you can’t even, say, substitute all A’s for G’s and vice versa and make resulting cell work without also redesigning all the proteins that bind to specific DNA sites. And substitution of subset of amino acids our cells use for another one would wreck everything…

You can still of course brute-forcely emulate every neuron (not being able to do so would be indeed indication of some mystic substance at work) but very computational power needed to do that is so baffling that even if Moore’s law would hold true for much longer than we can imagined, by the time it’s developed it may become irrelevant anyway (say, another, simplier technique for extending lifespan appears). And such a brute-force simulation would be difficult to “improve”, so it still may become sick, may even have to age and die to work…

Posted by on 12/04 at 09:19 AM
Sorry for any bad English, by the way.

Also, the very ability to “scan” the brain would make “brute-force upload” useless as a research tool. So, “brute-force upload” is most probably patently useless, even if possible.

Posted by on 12/04 at 09:44 AM
Hi Anonymous,

I just edited your comment to restore the link to the Intercal Wikipedia page.

I am not completely sure I understand the point you are making. I think your point is that a computational mind can only be achieved with the help of (very) high level models and languages (that is, based on high level functions rather than low level “assembly code"). Correct?

I probably agree, but for the task of “copying” a person to another computational substrate where she can continue to function as a program implemented in this high level language, you may still need the ability to interface to low level microscale (and even nanoscale) physical models in order to get the data in and translate it to the high level language.

Posted by on 12/04 at 10:42 AM
I just edited your comment to restore the link to the Intercal Wikipedia page.

Thanks.

I am not completely sure I understand the point you are making. I think your point is that a computational mind can only be achieved with the help of (very) high level models and languages (that is, based on high level functions rather than low level “assembly code"). Correct?

Well, any practically useful re-implementation ought to be “high-level”, or you simply won’t be able to do anything interesting with it.
Yet, there are few important consequences:
1) As our cells indicate we probably won’t find such a high-level language by mere studies of brain as it is.
2) Need to run nano-scale simulations for even parts of the brain cranks up computing power requirements to a sky-high level. (Look, that IBM supercomputer isn’t as compact or energy-efficient as a natural rat. And it would spend years trying to computationally fold a protein (I mean ab initio folding, of course, not one using correlations between known proteins.) )

3) (1) and (2) basially mean that well, we can’t say anything about practical mind uploading yet. It may be in fact useless anyway, for some reason we don’t know. (Like atomic pen . We can make one, but no one would want it.) Or even practically impossible (again, ability to do “brute-force” upload, not “high-level” upload is enough to prove brains materiality. And brute-force upload is almost certainly useless.)

Posted by on 12/04 at 11:34 AM
The blog posts don’t actually talk about Dawkins (they talk about “Universalism” instead). “Universalism” is basically what you are talking about on Amor Mundi.

Posted by on 12/04 at 05:23 PM