The author seems to have written the article on the assumption that it will be humans figuring out a brain upload mechanism.<p>While his time frame estimates might be sound given that context, I'd argue they're wildly inaccurate given a more likely scenario: such things will be brought into existence first via a super-intelligent agent, if at all.<p>If that proves to be the case, the time frame for brain uploading becomes roughly bound to the advent of AGI. When that happens, the same entity capable of creating our brain upload mechanism would likely be capable of endowing humans with effectively immortal bodies just the same.<p>In other words, technology like this is almost certainly going to be part of a post-singularity world (assuming there is a singularity in the first place), and who knows what that will look like.
I think discussions around this are ill-served by the "upload" metaphor, as it somewhat implies that the new brain, the original brain, and the uploading process are all largely separate things. It seems much more likely to me that the new brain will take the form of an augmentation that survives the death of the original brain. The personality would be "uploaded" not by deliberately sampling parameters to feed a model, but by the new brain organically (heh) becoming a component and embodiment of the existing personality. I'm sure this isn't a new idea, but I don't know what it's called.
When I look at efforts to synthesize the human brain within a computer I see an effort to make a synthetic bird rather than a machine that flies.<p>If this analogy has any use as a model beyond a cursory observation then a focus on machines that solve hard problems will bear more fruit and earlier (maybe ever) than attempts to fully replicate the human brain.<p>Today is it possible to build an (outwardly) anatomically correct and functioning sparrow (in all respects) out of synthetic materials? I don't know but surely it is hard. To make one certainly would have been impossible for the Wright brothers.<p>Today we have things like the F22, A380, and little drones. Each is quite complex but the complexity accreted over time, each layer a pragmatic solution to a problem at hand.<p>If we take the same approach what kind of "thinking machine" might we end up with in 100 years time?
> We all find our own solutions to the problem death poses.<p>No! No body does this. People make up excuses. People spin death to be a positive, often under some guise of Deep Wisdom. People come up with all sorts of ways to <i>cope</i> with death, but calling any of them a solution is just false. Death is a vile, atrocious thing, the biggest enemy humanity has.<p>We don't say that slaves all found solutions to slavery. Or that everyone finds solutions to domestic abuse. Or solutions to dementia or Alzheimer's. Death is a far greater evil.[1] So how disgusting is it to say everyone finds a solution.<p>I admit the author probably didn't intend to imply this, but it's exactly that kind of thinking that we should be aware of and fight. Just because it seems inevitable, we should not make it socially acceptable to give up and view death as anything but the wickedness it is.<p>1: Yes there are atrocities worse than death, but many of those involve death, or are worse because of the limited timespans caused by death. Apart from having your mind destroyed, I'm guessing most things would be healed by rather long periods of time, than sufferers would prefer a period of suffering+long OK life, vs suffering+death.
So, my concern about uploading my brain, is whose cloud service do you trust to run your consciousness? Google? Microsoft? Amazon? Facebook? Apple?<p>The level of trust that I'd require is pretty high. You could imagine it once again being relevant to check the pedigree of a company. Ideally there'd be a cloud provider already running now that is known for its trustworthiness. 'established 2013' might actually mean something one day.<p>My main hope is for indistinguishability encryption to reach the level where I wouldn't really have to trust the provider, but as long as you pay a significant performance penalty you're going to end up massively disadvantaged.<p><a href="https://www.youtube.com/watch?v=IFe9wiDfb0E" rel="nofollow">https://www.youtube.com/watch?v=IFe9wiDfb0E</a>
'When the construct laughed, it came through as something else, not laughter, but a stab of cold down Case's spine. `Do me a favor, boy.'
`What's that, Dix?'
`This scam of yours, when it's over, you erase this goddam thing.'
While I have no hard evidence for this, I expect we will eventually find that human intelligence and consciousness depends heavily on quantum effects. Thus it will always be impossible to scan and upload a human brain in a way that captures the essence of a person's mind.<p>Even though I don't think it will ever happen I enjoyed reading the hard science fiction novel "Hegemony" by Mark Kalina. It presents an interesting vision of what life would be like in the far future with mind uploading.<p><a href="http://www.projectrho.com/public_html/rocket/atomicnovel.php" rel="nofollow">http://www.projectrho.com/public_html/rocket/atomicnovel.php</a>
So, to justify their "just accept death, guys" conclusion, the author makes broad sweeping statements like "details quite likely far beyond what any method today could preserve in a dead brain" without providing any specific arguments about the limitations of vitrification whatsoever. Not even a bald assertion that "cryonics fails to preserve structure X". Yup. Okay.
Assuming that the ability to upload our entire brain to a computer will not be available during out current lifetime, what about the idea of mass data collection of our experiences, storage of this data, then periodically run improving algorithms that combine all our experiences into a consciousness?<p>This would maybe involve wearing a camera 24/7 to record everything we see and hear, and also some feedback on our own inner thoughts and our own recording of our emotional responses to things. When we finally die, a computer crunches all this data using neural networks to create a consciousness based on our life experiences and emotional responses.<p>This would initially be maybe not fantastic, but as technology progresses the crunching of the source data improves and every decade a new iteration of our consciousness can be produced, hopefully coming closer and closer to the real us.
Permutation City by Greg Egan explores what it would mean if you could easily upload your consciousness but wealth meant access to better, faster, hardware to store it.
An additional complication is the role of the billions of glial cells that are present in the brain. Their function of supporting neurons is quite well characterised, but they can also more specifically modulate neuronal function, so their place in a comprehensive connectome model shouldn't be ignored.<p>Then there is the network of vasculature, the flow of cerebrospinal fluid, interactions with hormones, and just generally all the interfacing with other bodily systems. All this would have to measured and modelled somehow too, in addition to all the billions of neurons.<p>I agree with the author, there's no way all this is going to solved any time soon.
What would I do with an uploadable version of my brain? I'd send it off to an accelerated university where it can learn at a rate far faster than I ever could. Plus, if we could upload a brain, then there's a good chance that we understand how to manipulate the state of an existing brain. So after my brain has graduated, we'll simply upload that virtual brain's state back into my skull! All this would take about 2-3 min tops? But somehow I imagine this will still cause university enrollment prices to climb. It's student loans all the way down.
A couple of quibbles with the article:<p>>While progress is swift, no one has any realistic estimate of how long it will take to arrive at brain-size connectomes. (My wild guess: centuries.)<p>It's not so hard to estimate. Just extrapolate progress on scanning, computing etc. About 2050 plus or minus a couple of decades. We had a 20um scan in 2013 and you'd probably want to get that down to 20nm for a connectome so if you assume resolution doubling every couple of years that would be about 2035.<p>(2013 scan: <a href="http://io9.com/see-the-first-ultra-high-resolution-3d-scan-of-the-ent-514395280" rel="nofollow">http://io9.com/see-the-first-ultra-high-resolution-3d-scan-o...</a><p>Images showing neural connections: <a href="http://book.bionumbers.org/how-big-is-a-synapse/" rel="nofollow">http://book.bionumbers.org/how-big-is-a-synapse/</a>)<p>Of course as the article points out a connectome misses a lot of chemical detail.<p>>It will almost certainly be a very long time before we can hope to preserve a brain in sufficient detail and for sufficient time that some civilization much farther in the future, perhaps thousands or even millions of years from now, might have the technological capacity to “upload” and recreate that individual’s mind.<p>Or quite possibly we can do it just now for $30k or so by sticking the body in liquid nitrogen. (<a href="http://www.cryonics.org/membership/" rel="nofollow">http://www.cryonics.org/membership/</a> ). Maybe that won't work but maybe it will.
With modern ML techniques it looks feasible to recreate at least online behavior of an individual - facebook likes, comments etc. The datasets are here, in facebook/google datcenters, and DL models are already used to model conversation.
It would be interesting to know just how many megabytes of logs of your online activities is really necessary to extrapolate your behavior into the future.<p>Facebook AI research is probably playing with such models right now.
TL:DR; The brain is very complex so it will take long, I don't know how long and I don't dare to make an estimate, but, as I'm getting older I am more and more at peace with dying.
I sense this will be one of those articles that will be a perfect example of someone who had a very firm position against progress is clearly proven wrong. I wonder if we could automatically flag articles like these based on the number of tautologically negative arguments that appear...
Upload my brain, so people in the future can use it as a DIY kit for Artificial Intelligence?<p>This article sounds a lot like SOMA the video game. brain scans, uploading your brain, etc
<a href="http://somagame.com/" rel="nofollow">http://somagame.com/</a>
i find it hard to imagine how to replicate the smaller connections of my brain, and the inter-connections that they all have.
With that being said, i hope that it could be done, because i would be super down, and would want to be uploaded into the net.
Great Mambo Chicken and The Transhuman Condition<p><a href="http://www.amazon.com/Great-Mambo-Chicken-Transhuman-Condition/dp/0201567512" rel="nofollow">http://www.amazon.com/Great-Mambo-Chicken-Transhuman-Conditi...</a><p>Very interesting book.
seems hardly worth the trouble. don't flatter yourself your brain needs to stay around - highly unlikely and very un-ecological to power up a machine to maintain a presence of virtualitzed shit for brains. -Just saying. Next Question?