Your mind will not be uploaded (2014)
softmachines.org
softmachines.org
But it might be enough like you that it could convince your bank to give it your money.
The brain upload isn't for power or personal immortality. It is for the thought patterns of the time to be accessible as a library or intelligent servant so that later generations may prosper.
If I am fortunate, I will be so preserved, and they will overrule me most of the time, using me the way I use (say) Karl Popper's books: as a reference, not as immutable truth, and not as the sole tool to make decisions. My mind is far more general purpose than a book, so it will be tempting for them to treat it as a sole tool, but I'm counting on them being smarter than me.
And if they deem me, Gen 1 uploadee, only worthy of jokes like we see Alicebot, so be it. If I am to be deleted as I am superseded, so be it.
> I can create copies enhanced by partner genetics and trained by me who have the ability to fit themselves to circumstances I cannot foresee.
Normal reproduction is good for passing on advantageous genetics and parenting is a short-term transfer of advantageous memetics. The brain upload is for advantageous non-degrading memetics. Think of the uploaded brain as just a book with an intelligent query engine.
In fact, I'd go so far as to say that if we're able to make servants of uploaded brains it will be as vast a leap in human capability as when we got the written text. We gained the technology to pass ideas across time and hit a hockey stick in growth. If we gain the technology to increase that fidelity and permit querying informed by additional information? Making 10 copies of me, with varying plasticity, and then feeding them information to see if they find novel solutions to problems?
Forget training a neural network to detect answers. Just emulate the brain and you get a never-tiring never-bored (since you can programmatically dopamine-train this being) eternal intelligent searcher who understands all your queries.
¹ Not strictly, since it's not a clone, but clones are shitty anyway since they're less capable of genetic adaptation.
Evolution is massively inefficient as a design process, and gives no guarantees about what you get.
Personal immortality is the goal.
I desire that death is not conquered because of that fear.
Sleeping doesn't generate petabytes of data elsewhere in space.
Sleeping doesn't have a mechanical "scanning" process which takes time to read across a brain.
Sleeping doesn't have a mechanical scanning process with measurement errors, approximations and simplifications.
Sleeping doesn't abruptly disconnect trillions of body cells.
Uploading doesn't use biology or Penrose and Hameroff's microtubules.
Uploading doesn't have a physical embodiement.
Uploading runs a simulation.
Uploading is a different place in space with a wildly different configuration of matter and energy.
Consciousness has not been proven to be a) purely information processing or b) possible on an electronic Von Neumann state machine. (I'm not saying it isn't, I'm saying "of course it is, what else could it be?" isn't proof - especially when a popular part of uploading arguments is the ability to simulate arbitrary things or fake parts of the experience to the point where you won't notice anything is wrong; taking a position where simulations can be OK because we can do that to make them be OK, but that can't be happening right now because nothing seems wrong, is weird).
You don't have to believe in a soul, quantum woo, or human exceptionalism to observe that part of the universe doing something and part of the universe doing something which looks to humans like doing something else, are different.
That's why I'm a fan of incremental substitution. Replace an individual neuron with an artificial one that duplicates its state and function, but is more mechanically robust. It's clear that doesn't matter to you — individual neurons even die without causing any problem, so an externally-indistinguishable substitute certainly won't. Now do that 100 billion times.
Let's turn it around: we start off with you and a perfect clone of you, except the clone has artificial neurons. Then we gradually swap your neurons with the clone's, without ever interrupting consciousness.
Using the logic of your thought-experiment, we should conclude that the 'real you' still resides in your original body, despite that all your neurons now reside in the clone's body. This seems unconvincing.
Perhaps the Buddhists have it right and identity of consciousness is illusory, so all such thought-experiments are meaningless.
Personally however I would view a perfect clone of me to be the same as me. If my brain were absolutely perfectly replaced in an instant of zero time then yes, I would consider that to still be me.
There's a great tabletop rpg that goes into this called eclipse phase. In it most people have uploaded their brain and can jump around in bodies due to digital brains being placed into everything between clones and nanobot swarms. This also allows them to make copies of their mind at will to do tasks and merge back together. I'm definitely a transhumanist because I would love to have such technology and would be happy calling all those copies just as valid as the original.
This is true as a practical matter, but it doesn't have much bearing on the thought-experiment. We can just add the additional constraint that you and the clone are both in identical operating theatres and your thoughts don't diverge. I don't see that it matters much either way though.
> I would view a perfect clone of me to be the same as me.
But it isn't. If it suffers, you do not. You might not even be aware of its existence.
> most people have uploaded their brain and can jump around in bodies due to digital brains being placed into everything between clones and nanobot swarms
Star Trek didn't go that far, but it had the 'transporter', which disintegrates you and then synthesises a copy at the other end. It's mentioned in the series that some people refuse to use them for this reason.
This is what I was getting at. In the pure thought experiment it's a lot easier to go along with saying that both the pre and post replacement entity is the same person. But people who are ok with that ideal are usually going to be more worried once you involve those real effects that diverge the actual experiment from the thought experiment. If the process takes time then there is time for divergence and that possibility can cause anxiety. If there is no time then it is simply a question of "If you got instantly replaced with an exact duplicate, are you still the same person?" which is meaningless, you would have no possible way of knowing it had happened. You should be more worried about sleep actually being death and the person who wakes up being a different person.
>> I would view a perfect clone of me to be the same as me.
>But it isn't. If it suffers, you do not. You might not even be aware of its existence.
Let me rephrase this. Yes the clone suffers and the original wouldn't be aware of it. But I consider both people to be the same person. This is not to mean they are some sort of shared consciousness.
Let's imagine you had a clone and an original stuck on an island with supplies to last 4 weeks. You know help will arrive in 8 weeks, so one of you needs to die to allow the other to use the supplies to last that long. If it were me in that scenario I'd just flip a coin and the one who wins gets to survive on the supplies. I consider both as valid as the other to the claim of being me.
That's fiction, of course, but what the characters experience is what I'd expect to happen in real life if a transporter did exist. If a structure (human) is destroyed and recreated identically at some other location, I don't see why that should be any different than moving the atoms over while preserving the structure.
And I don't see why things would be different if you created the same structure (or an isomorphic one) inside a computer.
Of course, I'm assuming that what makes you you is simply the structure and how it operates as a system. That's definitely a philosophical question, but I think if its answer is what I expect, then a perfect recreation of you is you.
And yes, this raises the question of whether there can be more than one of you. (Which I think the Star Trek novel also covers.) I don't think this cloning presents a philosophical barrier, though. I think it's similar to calling fork(). The only observable differences come from the kernel intervening to change the pid and fork()'s return value. There's nothing fundamentally unsound about the idea of splitting into two and both being "you". Although there are definitely things that are unsettling about it.
Well, the people going in the transporter would be killed, so they wouldn't be feeling anything after that. The people coming out of the transporter might feel like they just appeared in a new place, but that is immaterial to the people who were killed by the transporter.
Similarly, a mother could have grounds for rejecting a teleported child, for the one who came out of the teleporter was not physically the one she gave birth to.
Incidentally, the philosophical concept that episode explores goes back long, long before Star Trek, to the Ship of Theseus[1] thought experiment from Ancient Greece (and I wouldn't be surprised if this was considered even before then in ancient India or China).
The Ship of Theseus problem asks: if a ship's parts are progressively replaced with new parts, until all the parts have been replaced, is it still the same ship?
The cells in our bodies are being progressively replaced. A materialist, naturalist, or physicalist would have trouble explaining how someone could retain their original identity if nothing of them was left after this process.
The brain upload, teleporter, and transporter questions are variations on these.
I don't think that consciousness has momentum.
You're really talking about (at best) a new instantiation of the identical consciousness in a new body. That's not the original consciousness moving, that's a new (identical) consciousness coming to life.
I am unconvinced a consciousness can migrate out of the body that it's in. Copies are not enough.
This is made more obvious if you consider the case that the original is not destroyed. Then the clone that comes out of the end of the transporter would be having experiences that the original can not experience. If they were really one consciousness they would have one experience from that point on, but they don't. They're two different people.
So in the first case one person dies and another person with the same body and memories is born, and in the second case the first person does not die and another person with the same body and memories is born. In both cases they are two separate people.
This is making a lot of assumptions about how consciousness works. If I make a copy of a book the story doesn't move, but it is now in two places. If I destroy the original book the story isn't lost. I think experience might be more like that. Multiple copies of the same information system will all only have one subjective experience.
Yes if you run two versions they will diverge and the question is much more complicated. But if there is one moment of copy/destruction or you're scanning and then simulating/rebuilding a frozen brain, then there's no point in time where a consciousness doesn't have a future version.
There's a mechanistic, reductionist point of view that runs throughout many of the HN comments here that I find really disturbing.
A lot of people seem to think that we are just machines. If we are, then sure, replacing one machine with another that works the same and has the same "data" might be unproblematic, but I don't share such a bleak and narrow view of humanity. To think and feel is not just to run a program on a computer.
I don't think it's less mechanistic to say that the second copy creates a second entity.
Why do you think that the thinking and feeling is tied to the original body?
Suppose I transfer a virtual machine from one physical server to another using some live migration tool. Of course it migrates the state of all running programs. But beyond that, for all practical purposes, the new VM assumes the identity of the old VM: its IP address, its pending requests, its role in a cluster, et cetera.
Someone might argue that the new VM is not really the same entity as the old VM; after all, it's stored on a different physical storage medium and processed by a different physical CPU.
I'd respond: who cares about the physical storage medium? That's the wrong layer of abstraction to be looking at. A VM does not exist on the plane of matter but on the plane of information. It is not made of hard disks and CPUs, but bits. The hardware is just a way to instantiate it in the physical world; it's not part of its identity.
So it is with people, even if we currently lack a way to transfer the bits representing them to different storage media.
The same is not the case with people.
It would matter to a lot of people if their loved ones were the ones they knew and had intimate contact with, not some clone of theirs that simply shares their appearance and memories.
The person to whom it would arguably matter most is the person getting destroyed or killed. What comfort is it to them that some clone of their lives on?
It has certainly processed events during its lifetime and accumulated data about them (~= experience that a person might get from living through events). There wouldn't be ethical issues with destroying it, but there would be the downside of losing that data; however, migrating it doesn't have that problem.
> The person to whom it would arguably matter most is the person getting destroyed or killed. What comfort is it to them that some clone of their lives on?
Personally I would have no problem with myself or anybody I know going through such a process. Partly because I would consider the result the same person, but partly because I think more broadly that the difference between "I live on" and "a clone of me lives on" is unimportant.
For example, suppose I went through a teleporter but due to some mishap the body at the sending end were not destroyed, so there were two copies of me. If this happened and were detected immediately, both copies of me would freely volunteer to kill themselves in favor of keeping the other one. From my perspective (as either of the copies), the only thing lost would be the memories I made since teleporting, but losing a few minutes' worth of memories is not a big deal and something that happens all the time. Beyond that, the other copy is still "me", so there's nothing to offend my desire for self-preservation. If on the other hand the situation lasted much longer before being detected, so I gained more unique memories, or if it was only a few minutes but I just happened to have some kind of eureka moment in that time, I might be more ambivalent.
Ideally there would be some way to merge the memories of the two copies into one person. In that case, I would be interested in intentionally making short-term forks of myself on a regular basis, to be merged later (say, after a few hours), so that I could effectively do multiple things at the same time.
Do you consider identical twins the same person? Do you consider Abby and Brittany Hensel[1] the same person - they share the same body so they must have shared the same life experiences?
> I think more broadly that the difference between "I live on" and "a clone of me lives on" is unimportant.
Do you think the difference between "I live on" and "my child lives on" is unimportant? What about the difference between "I live on" and "someone else lives on"? The clone is a different body in a different place, it's the same situation as "I live on" vs "any other human lives on", except for what you believe about the other person.
> If this happened and were detected immediately, both copies of me would freely volunteer to kill themselves in favor of keeping the other one.
Then why don't you suicide right now, if continuity of experience is that casually irrelevant to you and all you need is a belief that some other body in the universe shares the same ideas you do? There are a lot of people on the planet right now, are you not convinced that someone, somewhere, shares enough of the important ideas with you already? Why not? Is your memory of a dentist's visit in third grade, or your particular grandmother's generic apple pie really that special?
In the "accidental clone" scenario, I wouldn't just be losing memories of a few minutes of experiences; I'd also be losing whatever delta to my personality the discarded clone went through in those minutes. However, in most cases I wouldn't care because that delta would be miniscule. As I mentioned, a potential problematic case is if I suddenly had a eureka moment in those few minutes that single-handedly changed my personality, but that's pretty unlikely.
With Abby and Brittany Hensel, on the other hand, they've had three decades to build up unique personalities, so despite sharing many of the same experiences they're definitely different people. (Plus, there are various sources of divergence between their experiences. For example, each of them will remember a different subset of their experiences, and in any argument between the two of them, each of them would experience being on a different side of the argument.)
Similarly, "I live on" and "my child live on" are very different things; a parent and child don't even share most of the same experiences, let alone personality.
> Then why don't you suicide right now, if continuity of experience is that casually irrelevant to you and all you need is a belief that some other body in the universe shares the same ideas you do? There are a lot of people on the planet right now, are you not convinced that someone, somewhere, shares enough of the important ideas with you already? Why not? Is your memory of a dentist's visit in third grade, or your particular grandmother's generic apple pie really that special?
I'm a special snowflake… just like everybody else. [ed:] I'm not important, but I'm unique; nobody is exactly like me.
Do I make the world much better with my existence? Not really. Hopefully I make it a little better; hopefully I don't make it worse; but I'm only one person and not a particularly famous one.
But I still want to live, and I want other people who are alive to keep living. Not as a means to some end, but as an ingrained human desire.
Yet I choose to interpret "I" as "my special-snowflakeness", as opposed to tying it to some physical object [ed: or continuity or linearity of experience, which I think are largely violated already when I fall asleep or lose memories]. An exact copy of me is the same special snowflake, at least at first, progressively diverging as the two copies start to have different experiences.
It just seems weird to me that someone would think "I don't matter if there's a backup of my data somewhere and they could bring back a copy of me if they wanted." or "There's already a copy of me out there right now, so that makes my own death ok."
To me it wouldn't matter in the slightest if there was a clone out there of me. As long as I want to go on living, it's me having future experiences that matters to me, not whether some clone of me has these experiences.
Something else that hasn't been discussed yet is whether the clone really does have all your experiences and really is identical to you. How would you really know? Would you trust some scientist's or some machine's evaluation? I don't think you could ever really know.
Of course, for me it would make no difference as I wouldn't want to die so some clone may life, but it might matter to those of you who are willing to sacrifice yourself for a clone. In that case think: how would you really know they were a clone?
Sure, you could quiz them regarding some memories that only you had, but you couldn't ask them about everything. As important, you can't know if their experience of the world matches yours, nor what their true beliefs are.
Some people are surprised to find their life partners aren't who they thought they were, after decades of living together.. but you probably won't have decades to get to know your clone before you voluntarily suicide. In fact, in the "teleporter" example and in the mind upload example I'm hearing people eager to suicide even before their clone even exists. So no verification at all is even possible.
That kind of attitude just boggles my mind.
> "I'm not important, but I'm unique; nobody is exactly like me."; "I still want to live"
I submit that if you were in the accidental clone scenario, you wouldn't be willing to die as you claim, but would instead feel "I still want to live" very strongly, and your belief that the clone is "exactly like you" would give way to "they just came out of a nano-assembler and they know it, they're only an approximation of me who knows how many things are subtly different inside, I deserve to live more strongly" or on the other side "I just stepped out of a nano-assembler, I'm new and fresh and they've been ageing for decades and they intended to die in that scanner, I deserve to live more strongly", or possibly "I'm willing to share everything I have and compromise, there's no need for any death here".
> [ed: or continuity or linearity of experience, which I think are largely violated already when I fall asleep or lose memories]
I'm not happy with the casual idea that continuity of experience is violated by sleep. If that were true, waking up somewhere "unexpected" would be impossible, instead it's very disorienting and scary and a trope of horror stories. We expect to wake up where we went to sleep, we expect any times of waking up in the night to be in the right places in the sleep timeline, and we expect to feel an amount of refreshed/relaxed/over-tired/too hot/aching/with a numb arm from lying on it/etc. depending on how long we slept for and how well we slept (which is something we are aware of and can report on), and we can experience waking up with answers or decisions to things we were thinking about the night before. Sleep is not 8 hours of non-event.
If my body is failing, I'd be happy to make a copy, even let it run before I die. Not sure I'd want to if it was a destructive copy (it killed me). But some would I suppose.
However, if your life has meaning beyond just work and responsibility, then I don't think it matters if there's a clone, as your experiences would presumably matter to you, and it wouldn't matter if some clone was having them instead.
I think you'd find a whole lot of people unwilling to use a Teleporter, even over multiple generations.
That's begging the question, isn't it? If the destroy-and-copy view is right, then the people who use the transporter are dead, and replaced by the transporter with people who have the memory of having safely used the transporter.
I agree, though, that after a while society would just adapt and no one would care whether they'll survive the transport.
They kind of glossed over the morality - a wasted episode really. Didn't say much about which was the 'real' Riker. Didn't seem to care.
There were then two people with common memories up to a point. There could be legal complications, probably already thoroughly explored. Does Starfleet have two flag officers now, with two service records? Yes. If he were married, does his wife now have two husbands? Probably. Kids? Two dads, now. Property? Held in common, until they negotiate. Etc.
Say you could do a full biology simulation, and you use that to essentially extend your brain like extra layers of computation. Then you slowly turn off parts of your original brain while adding more chunks of functionality, until all of the original brain is gone.
This would be you in the sense that there is a continuous chain of being between your original self and your digital self (counting sleeping and maybe coma as still part of that chain of being)
Of course if this was enough to designate you still being you, it would throw into question if you stop your heart and then restart it, are you still you? (Probably the nature of stopping hearts is more complicated than I think it is, so I am not 100% sure this is a valid objection)
You don't know it (pun intended).
There is, in fact, no definition of what "you" even is. It's called a "hard problem of consciousness" for a reason.
But if you were standing next to a machine you had "uploaded" to, and I was asked to identify which was really you, I would have no slightest doubt. And, really, neither would you.
Which me are you asking? The meatbag and the machine would both claim to be me, and both would have some merit to that claim by having the same memories, same personality, etc.
After the uploading they'd start to go separate ways and develop their own exclusive memories. Initially the meatbag would probably be closest to current-as-I-am-writing-this-me, if that means anything, but the machine could take that title when the meatbag develops alzheimers or passes away.
You, of course.
Both would claim to be me, and I think both would be right. You're trying to force a certain conclusion by only asking one of them and by limiting there to only being one "real me".
It has a very pointed reference to the potential horrors of having your mind uploaded by someone who is trying to control costs.
I respect the work enough (I expect it's even harder to make money on short films than feature films) that I didn't bother tracking down any free copies. I know some people will, but I don't have to help.
I saw it on Netflix, but I doubt it's still there.
The child's doodles make it seem like a lighthearted thing. It is not.
A grownup is attempting to explain some fairly dark things to a very young child. Mercifully, the child cannot absorb what it is being told. You may not be so lucky. But there is some biting social commentary and the absurdity of the whole thing makes it work.
Making it more complicated for no good reason is just draining the fun out of it.
https://www.theverge.com/2020/1/22/21076806/google-janelia-f...
First the fruit fry brain is very small. You can image the entire thing at the microscopic level with a single image. The human brain is massive by comparison. Getting a coherent image that traces an axon from the tip of frontal lobe to the back off the occipital lobe is going to be a huge challenge.
Second the fruit-fly brain has 25,000 neurons while the human brain has more than 10,000,000,000. There's 6 orders of magnitude difference there.
Third, it's highly likely that glia (non-neurons) in the brain play a major role in neural computation so we'll have to image those too. Humans have way more glia than most other animals.
Lastly the connectivity of neurons in the human brain is very high. Getting those little connections right is key in all this as we aren't just going for the neurons but the connections between them.
>Lastly the connectivity of neurons in the human brain is very high.
yep, the human brain is 100B neurons and has 10e4 connections per neuron while the fly brain has 10e3 connections/neuron. So we need to emulate 10e15 connections of the human brain. GPT-3 has 175B of weights.
If you turn a computer off, then try to transfer the software it was running, you will get some data. But whatever was in the volatile RAM, won't get transferred. It's very likely the mind has such data.
Even drastic measures like electrical shocks or chemicals (up to a point) tend to have temporary rather than permanent effects. That evidence seems to imply that most of what we consider as 'us' is the more permanent physical neuron connections rather than the transient chemical/electrical states.
It's a challenge for reading the neurons connectivity for sure, but I don't think it is evidence that there is more to 'us' than our physical neuron connectivity graph.
"Ion channel proteins change shape in response to the electric field across the membrane, opening or closing pores; at the synapse shape-changing proteins respond to electrical changes to trigger the bursting open of synaptic vesicles to release the neurotransmitters, which themselves bind to protein receptors to transmit their signal, and complicated sequences of protein shape changes underlie the signalling networks that strengthen and weaken synaptic responses to make memory, remodelling the connections between neurons."
Can the weight of a connection be surmised after oxygen deprivation? Or are the chemical changes that happen under oxygen deprivation irreversible (from an information theoretic viewpoint)?
Well, really, there's no proof to that hypothesis. And I find the argument weak. The complex processes necessary for cells to live might have nothing or just a tiny bit to do with our consciousness.