Editor's Blog
OK, brain uploading fans. Defend the cause against this one…
six reasons why you’ll never upload your mind into a computer
“3. Insane Energy Demand
“The human brain only needs 20 watts to run the app called You, but with almost 7 billion of us and counting, we’re already straining the earth’s ability to host us all. Meanwhile, you know how much juice one Google data center consumes just to index the latest LOLcats (a task much, much simpler than hosting your digital consciousness)? 100 million watts. Do the math: We’d have to invent fusion reactors or build a Dyson sphere just to keep the lights on. Neither of those technologies are theoretically impossible — in fact, they fit right into the Singularitarians’ techno-magical worldview. But they’re definitely not gonna happen within the next few decades, and probably not even within the next century or two.”
Sign up for the Humanity+ newsletter:
13 Comments
If this were an argument against mind uploading EVER happening, I would point out that it must be possible to design something capable of doing what the brain does, without dangerously overheating. We know this through existence-proof: The brain itself performs this very trick. Researchers like Henry Markram hope that, by reverse-engineering the circuitry and functions of the cortical column and neocortex, we will discover ways to build computers/software closer to the capabilties that nature has shown to be possible.
So, in one sense this argument is akin to somebody in the 1950s dismissing 1ghz laptop computers, because vacuum tube-based computers would have to be the size of Manhatten and would catastrophically overheat and they need teams of ace engineers to keep them running. So forget userfriendly little boxes that cost a few hundred bucks. Arguments like that are not relevant to Integrated Circuit-based computers.
But, then again, this argument seems to be against NEAR-TERM uploading. How close are we to making the paradigm shift to an entirely new generation of computers (such three dimensional molecular computing, which Kurzweil champions as the 6th paradigm to replace the integrated circuit), which might address this particular problem? It may be that I am guilty of linear thinking, but I do not see mind uploading as a near-term (say, by 2030) possibility.
#1 Fail whale: Argument is “we can’t do it now, so we will never be able too” Software and hardware stability are much less of an issue now then they were even 10 years ago. By the time uploading is feasible, I expect much more robust systems to be in place.
#2 Storage Media: Same argument as above. Media longevity is not an issue so long as accurate transferral of data can take place. New media in the labs does not share the same “limited life” as magnetic or CD storage.
#3 Insane energy demand: Same argument as above. We are already shifting towards graphene and other computing technologies which use far less energy to perform far more “work” A graphene computer running in “particle” mode electronics (i.e. electrons as discrete particles) could run 1000x faster than the fastest chip made of silicon, while using no more energy. A graphene computer running in plasmionic mode (sp?) (i.e. electron as a “quantum wave” function) could run a thosand times faster than that. We are also likely to have energy sources many orders of magnitude more efficient and powerful.
#4 Lack of computing power: Same argument as above (can’t they come up with a new one?) See my answer to #3
#5 Brains need Bodies: (Athena Andreadis’s favorite argument.) Bodies and minds are interactive and inseparable. A variant on the “Soul” argument. Argues exceptionalism of the biological form. Is based on assumption that a “body” will not be available in either VR or Physical form. Current research already shows the mind can “self identify” with a virtual character when a persons “sense of place” is centered on the virtual character. Also, with the amazing advances in cybernetics, the likelihood of “telepresence” robots able to act as “hosts” to uploads in non-virtual settings seems high. Basically, while there may be a need for a “body” the definition of what constitutes one appears to be fluid, with the brain willing to accept alternate forms of “input”
#6 It’ll cause a war between have’s and have not’s! This one is so cliched it’s almost not worth addressing, but I will anyway. It will, however, take a reference to a recent blog post I made: http://valkyrieice.blogspot.com/2010/07/on-government.html
In a nutshell, this problem is likely to be addressed far sooner than uploading is likely to feasible. To understand the reasons why I say this you need to understand the root problems as I outlined them in the above blog post. A hundred thousand other technological developments in the works will “level” the field, and change the very definitions of “wealth”. Uploading is unlikely to occur until AFTER this has happened, making the likelihood of “restrictions” very small.
There are plenty of researchers who are developing neuromorphic hardware. Neuromorphic chips are different from conventional CPUs as they are specifically designed to model neural networks. This is analogous to how most games use customized GPUs to do stuff that would be impossible using an all-purpose conventional processor. Neurogrid is one project that will simulate 1 million neurons using one watt. The facets, spinnaker, and DARPA SYNAPSE projects are all attempting to do something similar. Memristors and carbon nanotubes are two alternative technologies that could help create embodied neural networks. There are various other methods that could also work. I suspect software running on regular processors would not be able to sustain qualia, embodied hardware on the other hand may be much more effective in attaining this goal.
At the very least, implants could be added to the brain to give us many of the benefits of hypothetical mind uploading scenarios (downloading information and different modes of consciousness, a programmable mind).
Has the author of these idiocies ever heard of the common sense idea of keeping the important stuff and discarding the useless stuff? Does he really mean that farts and diarrhea (and cancer) are as much a defining part of our identity as our thoughts and feelings?
I think the arrival time of mind uploading depends on how you define it. I believe we will be able to store bits of our minds in different mediums first and this will gradually grow, as will the real time connection and communication between the non-bio substrate mind and that existing on neurons.
I think your instinct about total mind upload not being available by 2030 is correct, but do I think that by then we will have partial mind uploading where our brains are in constant communication with the non-bio parts of the mind and I would definitely call this mind uploading, downloading too for that manner, since the communication between the bio and non-bio would be back and forth.
I’m thinking about the advances in chips being developed for Parkinson’s sufferers and extrapolating from there. They already have a primitive two way communication:
http://www.sciencedaily.com/releases/2009/11/091111111301.htm
To be fair, no farts, no fart jokes. Is this not a founding artifact of human culture and civilization? We must at least model these functions to the highest level of resolution possible!
@ Valkyrie Ice: The article was given the title, ’6 reasons why you’ll never upload your mind to a computer’, which certainly gives the impression that the author rules out the possibility of mind uploading altogether. However, in the article itself he does include the caveat, ‘Now you may be thinking, “Listen, jerk, you’re just talking in terms of today’s technology — not the untold wonders that will be invented before any of this stuff happens.” And you’d be right’. I do not think the author necessarily considers these 6 reasons to be ABSOLUTE barriers to achieving mind uploading. Rather, he considers them reason to believe uploading is not a near-term prospect. The title, perhaps, should be read as ’6 reasons why YOU will never upload…’, because the technology lies many generations into the future, if at all.
@ Giulio: I do not think this person is an idiot. He has raised legitimate concerns. Yes, IF we continue to solve the software complexity problem and IF we can radically improve the energy efficiency of our computers and IF LAR continues long enough to ensure the hardware capabilities are achievable and afordable and IF minds can be copied over to a new substrate with no loss of personal identity (and probably many more IFs)……THEN mind uploading is possible. Frankly, at this point in time it is a matter of faith to say that every one of this IFs will be addressed any time soon. True, Ray Kurzweil’s ‘The Singularity Is Near’ shows that there is ongoing R+D aimed at overcoming all of these obstacles (though not all this research is aimed specifically at achieving mind uploading) and that we know of no physical law that rules out brainlike machines capable of supporting our self-consciousness (indeed, if such a law existed our brains would be impossible), and it is also a matter of faith to declare these problems unsolvable. But they are enough to give those who think mind uploading is a near-term prospect pause for thought.
And since you’re not totally dumb, your distortions are clearly a deliberate attempt to mislead.
What I actually said was that human minds as they have evolved are inseparable from their brains and their bodies, insofar as the bodies provide sensory and motor feedback to the brains/minds. If you think even superficially, you will realize that this argument is the opposite of a separable soul. I also said that non-human minds may operate differently.
But far be it from me to deprive you of the comfort of endlessly repeating your religious mantras.
*sigh*
Your argument, like that of the article, argues a “uniqueness” that is “impossible” to duplicate about the human body and mind, claiming that since this “unique element” is “uncopyable” that it makes uploading impossible.
As such, it is IDENTICAL in form to the “Soul” argument, which also claims that a “Unique Uncopyable Element” would prevent “uploading”
They use a metaphysical excuse, you use a psuedoscientific one. But it’s still the same argument.
X is unique, which means X is uncopyable. Therefore Uploading cannot be done since X would prevent it.
Changing what X stands for doesn’t change the form of the argument.
I’m coming in late so spare me if I’m missing your point.
Inseparable in what fashion? The input/output bandwidth, if you will, is primarily channeled through our spinal chord. As the countless surviving paraplegics have shown, not every input/output function is necessary for our brains to function.
There are likely portions of that mind that are dedicated to these areas, but if we can download/upload thoughts and memories there is no reason we couldn’t emulate these inputs if they were even required.
- Distaste for blade cores in the cloud, and/or administering awesome body through it, that overmatches and continues distaste for body.
- Robot perfectly cheery good at cleaning the kids’ bathrooms, before Mind uploaded into it.
- Body-pillow cases with Suzumiya Haruhi on them no longer scandalous/fun.
- Meatpuppets hoarding the good camera rigs.
- VM sensory interface consistency means Haifa looks like Halifax and Harbin, and feeling the scenes there through someone else’s bones shrinkwraps the feeling a bit.
- Velocity-sensitive keyboards just got ubiquitous in education, but digital people play them like overgrown SID chips because the acceptable training time keeps getting flipped around.
- Licensing virtual milk, gravity soma means farting is still subsidized.
- Singularity cafeteria cilantro soma aren’t the varieties you hoped for.
- [4200 AMA competitors] # [Hosting hiccups]
- Not apostate, thank you.
Your reply to Giulio is misleading. You’ve suggested that certain ‘ifs’ are required to be fulfilled before uploading the mind is possible. However, I feel that we can all agree that uploading the mind may be possible whether or not it is immediately feasible for wide-scale use, practical, ethical, or any of these other arguments that are completely subjective with regards to whether or not it is possible.
i.e. We may someday realize that, indeed, it is possible to upload information from our minds to a digital format but that it takes a long time to do. The time factor simply makes it difficult to put into a business model. It does not eve imply that uploading the mind is impossible.
By the way, most of these issues are not even actual issues but ‘straw man’ arguments. “Uploading the brain is impossible because it might take too much bandwidth”
Well, does it or doesn’t it? Neither the author or yourself have the answer to that question and thereby this is simply not even a valid argument against the possibility that uploading our minds is economically feasible but rather something we should consider while engineering the devices to do so.
Those points have been made to her before. In all likelihood, she’ll ignore you.
Athena is a “anti-transhumanist” She doesn’t believe that the human body can be improved upon, she doesn’t believe cryonics can work (even if they do bring the corpse back to life, it won’t be “You” because “You” died) That uploading won’t work (because it’s just a copy, even if you slowly transfer on neuron at a time, once your last biological neuron is gone, you’re dead) and she considers the idea of “indefinite life extension” as either impossible, or an anathema.
She vehemently denies it, but as I pointed out, most of her objections come down to either “we can’t do it now, so it will never be done” or “Humans have Unique Essential Element X, which cannot be defined, copied, transferred, or restored once the biological body has ceased functioning.”
But please, don’t take my word for it. I recommend looking up all her articles and commentary on this site and making your own decisions. She is extremely knowledgeable about bioscience, and if you can get around the “I said so, so it is absolute fact” rhetoric, she does an excellent job defining the areas in which additional research is needed, and a very good job explaining the current “difficulties” that must be overcome.
Just remember Clark’s 1st Law: When a distinguished but elderly scientist states that something is possible, she is almost certainly right. When she states that something is impossible, she is probably wrong.