There are people who have traumatic brain injuries who could benefit from this research not in a distant possible future but right now. Some people alive today have only limited and fractured access to their memories. Better external memory capture and retrieval might be of value to them. I do not suggest ignoring one group in favor of the other, but suggest applying these ideas now will improve their application later.
My first thoughts on any technology are 'how can this be of use to the disabled, and in search & rescue?' If it can of use to either or both, the benefits often roll out to general use. In this case, all the way to a future 1:1 model of myself.
Everybody could use their memoirs and photo as external memory. Aging kills memory and external memory could help.
Part of my motivation for creating YouTube econ educational videos is that it might someday help a superintelligence bring me back.
The more idiosyncratically wrong the material in those videos is (and the less they just present economists' consensus in a straightforwardly clear manner), the more effective they will be for that goal. Mmmm, incentives.
You're right, and have given me the perfect excuse to justify any so-called mistakes I made in the videos.
I question whether future society would be willing to bring someone back to life even if it was clear that the person wanted to be brought back and there was sufficient information stored to allow it to happen.
There might not be a moral reason to bring someone back to life, because if future agents value content agents, they presumably would be able to create far more content agents far more easily by engineering agents from scratch for maximum contentment with minimum resource use.
There might not be an economic reason to bring someone back to life, because future agents would be able to make far more efficient workers than a 21st century humans.
There might not be selfish reasons to bring someone back to life, because though having a 21st century person in a far more advanced world might be interesting, I think superintelligences could find far more interesting things to do.
Perhaps one way to increase the probability of being brought back to life is to set up a trust fund or something of the like to do so. Some cryonics organizations have trust funds for bringing back their preserved bodies, which is similar. What do you think of this?
Best known plan for implementing digital immortality:
Initial stage – 1 month
I think it is a basis which could provide 90 per cent of needed information for Digital immortality. One can add other more sophisticated techniques later. Most of them are listed in the digital immortality map.
Second stage – 1 month
The third part should include all other techniques listed in the digital immortality map. Also investments in preservation of the information, like webarchives, youtube uploading, M-disks.
As seen in the first episode series Caprica, quoth Zoe Graystone:
"(...) the information being held in our heads is available in other databases. People leave more than footprints as they travel through life; medical scans, dna profiles, psych evaluations, school records, emails, recording, video, audio, cat scans, genetic typing, synaptic records, security cameras, test results, shopping records, talent shows, ball games, traffic tickets, restaurant bills, phone records, music lists, movie tickets, tv shows... even prescriptions for birth control."
I, for one, think that the meme-mix defining our identity in itself could capture (predict) our behavior in large parts, foregoing biographical minutiae. Bonesaw in Worm didn't need precise memories to recreate the Slaughterhouse Nine clones.
Many think we can zoom out from atoms to a connectome, why not zoom out from a connectome to the memes it implements?
Bonesaw in Worm didn't need precise memories to recreate the Slaughterhouse Nine clones.
But as it is said, do not generalize from fictional evidence.
The main question is should we consciously write down our secret thoughts and child memories, hoping on better reconstruction in the future? If some kind of reconstruction is inevitable, if AI will make some kind of simulations anyway, maybe it is better to provide it with as much correct information as possible?
I know at least in our specific community, that we'd rather be resurrected than not, and especially in a techno-utopian future, almost goes without saying, but it still worries me that you don't seem to mention consent. At least the top paragraph suggests a third party collecting information about someone else so that they can be resurrected after their death, and even if we skip over the more normal issues with doing that, resurrecting someone without their permission seems like a violation.
In the mix with the problems you've listed under 1. is whether this kind of resurrection is even necessary. Personally, I doubt those identity problems can be conclusively solved even in principle, at least to a level that people's intuitions don't dominate, although I'm inclined to give up on what's actually factual there and try to convince myself of the weakest notion of identity I can find believable. I can't do much pushing there, but the notion I default to using based on my intuitions (sleeping doesn't kill you, uploading does kill you) is hard to justify so I don't mind trying to push away from it. Should really be stepping up my DI efforts.
Basically we do it all the time when we communicate with a person and create his model in our head. I think that it is moral to return to life everybody, except whose who explicitly and rationally were against.
We don't know how to solve identity problem now, but maybe we will do some kind practical research and we will find it in the future. Or may be AI will help us. Until that I suggest conservative approach to identity - try to preserve as much as possible and accept copy creation only if alternative is death.
May be we could build mechanism of identity transfer which is independent from information. If identity has any substance, like soul or causal links, we could build machines that find it and preserve it.
There is also two type of immortality. Immortality for me, that is immortality from the point of view of the observer, which is most interesting, but also immortality-for-others, thats is immortality for your friends. Big world immortality from the link may work, but only for immortality-азк-me, but not for my friends who may want to see me alive in 20 years here on Earth.
Also big world immortality helps cryonics and DI because resurrected DI and cryo client will dominate big-world resurection landscape and some of these resurrections will be exact as originals. So big world immortality help to fill gap lost during cryo
Storing data that might be used to reconstruct someone in the future isn't really objectionable, but that seems separate from actually using that data to create the resurrection. And it probably works out fine in the utilitarian calculus unless you count the sunk cost vs creating a "better" new person or a utility monster, but bringing someone back to life just because they didn't mention that they didn't want it, or you thought the reason they gave for not wanting it was irrational, sounds really skeevy. We have rules about consent for interacting with other people's bodies, I think that includes implanting their consciousness in new bodies.
Most people hadn't chance to rationally evolute would they want to be resurrected and especially by the means of DI. Of course we could model their answer, but we have to create their model first which create circular logic in this case. many religious people probably will prefer not be resurrected by DI, because they bet that they could get better type of immortality in the other world. Also a person whose body is cryopreserved would rationally prefer be resurrected based on this body, but not using DI. It is large open field of ethical and legal questions here. What if a child died, and his father wants his DI (+DNA) immortality and his mother doesn't?
How can we solve the "heap paradox"?
Even if we assume that if you have access to all information about a person's every molecule and every signal in their brain allows you to resurrect the same person instead of creating a new person with the same memories (an assumption not everyone agrees with), where can we draw the threshold between resurrecting the person and creating a new person with some passing resemblance, if the amount of information becomes smaller and smaller? When does it stop being the same person and when can we speak about a completely different person with only a few similarities to the deceased one?
We could solve it if we move from binary idea of identity, which only can be 0 or 1, me or not me, to continuous idea of identity, which could have many gradation. In this case we should ask how similar the copy will be to the original or what is the probability that I find my self in the next moment to be that copy. The main problems with identity is that it is wrong notion. It assumes that set of all possible observer is divided on many subset of observers, which are identical inside that subsets. But the idea identity is only useful to answer questions like "what I will feel in next moment". Most such question could be answered without binary conception of identity. Any way I suggest ti stick for now to conservative approach to identity problem - see comment above.
In case of DI, we should collect as much information as possible.
The evidence provided of any dead person produces a distribution on human brains, given enough computation. The more evidence there is, the more focused the distribution. Given post-scarcity, the FAI could simply produce many samples on each distribution.
This is certainly a clever way of producing mind-neighbors. I find problems with these sorts of schemes for resurrection, though. Socioeconomic privilege, tragedy of the commons, and data rot, to be precise.
It could be solved by acasual trading between parallel worlds. I tried to explain it in the map under the title that DI stalks well with many world immortality.
If we have infinitely many worlds with the same evidence about the person ( but the person is different in different world, only the evidence is the same), we could create only one resurrection in each world which is in agreement with this evidence, AND it will be exact resurrection of the person from another world.
BUT, the person from this world will be exact resurrected in the another world, so each person will have exact resurrection in some world, and each world will have only one person which much its evidence. (So, no problems with ethics and resources.)
I think that it may be difficult to explain this in several lines, but I hope you grasp the idea. If it is not clear I could try better explanation.
If we don't know 100 bits of information, we need to create 2 power 100 copies to fill all gaps. Even for FAI it may be difficult. Also it may be unpleasant to the copies themselves, as it would delude their value to outside world.
If a gap is about very important feature or a secret event, it could be two completely different people. Like if we don't know if a person of interest was a gay.
...Okay? One in ten sampled individuals will be gay. You can do that. Does it really matter when you're resurrecting the dead?
Your own proposal is to only sample one, and call the inaccuracy "acausal trade," which isn't even necessary in this case. The AI is missing 100 bits. You're already admitting many-worlds. So the AI can simply draw those 100 bits out of quantum randomness, and in each Everett branch, there will be a different individual. The incorrect ones you could call "acausal travelers," even though you're just wrong. There will still be the "correct" individual, the exact descendant of this reality's instance, in one of the Everett branches. The fact that it is "correct" doesn't even matter, there is only ever "close enough," but the "correct" one is there.
I think that there is 3 option in case of incomplete information.
I suggest to choose variant 2. In this case everybody is happy. The subject is almost exactly resurrected in one of the universes. Each universe get a person which corresponds its conditions and do not get useless semi-copies of the subject.
Resurrect one individual, filling gaps with random quantum noise.
Resurrect all possible individuals with all combinations of noise.
That is a false trichotomy. You're perfectly capable of deciding to resurrect some sparse coverage of the distribution, and those differences are not useless. In addition, "the subject is almost exactly resurrected in one of the universes" is true of both two and three, and you don't have to refer to spooky alternate histories to do it in the first place.
So, as I understood you, you stay for resurrecting of "sparse coverage of the distribution", which will help to prevent exponential explosion of number of copies, but will cover most peculiar of possible copies landscape?
While I can support this case, I see the following problem: For example, I have a partner X, which will better preserved via cryonics, but my information will be partly lost. If there will be created 1000 semi-copies of me to cover the distribution, 999 of them will be without partner X, and partner X also will suffer because ve will now care for other my copies. (Ve could also be copied, but it would require coping of all world).
If it were my choice, I prefer to lose some of my memories or personal traits than to live in the world with many my copies.
(Ve could also be copied, but it would require coping of all world).
Why would that be the case? And if it were the case, why would that be a problem?
I think that there is 3 option in case of incomplete information.
I suggest to choose variant 2. In this case everybody is happy. The subject is almost exactly resurrected in one of the universes. Each universe get a person which corresponds its conditions and do not get useless semi-copies of the subject.
Scott Adams describes how he is already doing what you describe as "Cheapest way to immortality" here.
Yesterday I agreed to fund a small project that will, as a side benefit, end up putting most (but not all) of my personality into digital form. Add to that the extensive history of my writing that exists on the Internet,
Digital immortality seems much cheaper than cryonics and of similar effectiveness. Why isn't it more popular?
It has less chances of success as there is several non solved theoretical problems of personal identity. Also there is no much public leaders approving it. Yudkowsky was against. In my plan is just plan C - if antiaging and cryonics fail.
Identity is evolving with age. It is only DNA for a newborn. But rational mind and especially AI is responsible for his identity and could control it. An AI could declare that any other AI with the same set of goal is him.
If someone has died it doesn’t mean that you should stop trying to return him to life. There is one clear thing that you should do (after cryonics): collect as much information about the person as possible, as well as store his DNA sample, and hope that future AI will return him to life based on this information.
Two meanings of “Digital immortality”
The term “Digital immortality” is often confused with the notion of mind uploading, as the end result is almost the same: a simulated brain in a computer. https://en.wikipedia.org/wiki/Digital_immortality
But here, by the term “Digital immortality” I mean reconstruction of the person based on his digital footprint and other traces by future AI after this person death.
Mind uploading in the future will happen while the original is still alive (or while the brain exists in a frozen state) and will be connected to a computer by some kind of sophisticated interface, or the brain will be scanned. It cannot be done currently.
On the other hand, reconstruction based on traces will be done by future AI. So we just need to leave enough traces and we could do it now.
But we don’t know how much traces are enough, so basically we should try to produce and preserve as many traces as possible. However, not all traces are equal in their predictive value. Some are almost random, and others are so common that they do not provide any new information about the person.
Cheapest way to immortality
Creating traces is an affordable way of reaching immortality. It could even be done for another person after his death, if we start to collect all possible information about him.
Basically I am surprised that people don’t do it all the time. It could be done in a simple form almost for free and in the background – just start a video recording app on your notebook, and record everything into shared folder connected with a free cloud. (Evocam program for Mac is excellent, and mail.ru provides up 100gb free).
But really good digital immortality require 2-3 month commitment for self-description with regular every year updates. It may also require maximum several thousand dollars investment in durable disks, DNA testing, videorecorders, and free time to do it.
I understand how to set up this process and could help anyone interested.
Identity
The idea of personal identity is outside the scope of this map. I have another map on this topic (now in draft), I assume that the problem of personal identity will be solved in the future. Perhaps we will prove that information only is enough to solve the problem, or we will find that continuity of consciousness, but we will be able to construct mechanisms to transfer this identity independently of information.
Digital immortality requires a very weak notion of identity. i.e. a model of behavior and thought processes is enough for an identity. This model may have some differences from the original, which I call “one night difference”, that is the typical difference between me-yesterday and me-today after one night's sleep. The meaningful part of this information has size from several megabytes to gigabits, but we may need to collect much more information as we can’t now extract meaningful part from random.
DI may also be based on even weaker notion of identity, that anyone who thinks that he is me, is me. Weaker notions of identity require less information to be preserved, and in last case it may be around 10K bytes (including name, indexical information and basic traits description)
But the question about the number of traces needed to create an almost exact model of a personality is still open. It also depends on predictive power of future AI: the stronger is AI, the less traces are enough.
Digital immortality is plan C in my Immortality Roadmap, where Plan A is life extension and Plan B is cryonics; it is not plan A, because it requires solving the identity problem plus the existence of powerful future AI.
Self-description
I created my first version of it in the year 1990 when I was 16, immediately after I had finished school. It included association tables, drawings and lists of all people known to me, as well as some art, memoires, audiorecordings and encyclopedia od everyday objects around me.
There are several approaches to achieving digital immortality. The most popular one is passive that is simply videorecording of everything you do.
My idea was that a person can actively describe himself from inside. He may find and declare the most important facts about himself. He may run specific tests that will reveal hidden levels of his mind and sub consciousness. He can write a diary and memoirs. That is why I called my digital immortality project “self-description”.
Structure of the map
This map consists of two parts: theoretical and practical. The theoretical part lists basic assumptions and several possible approaches to reconstructing an individual, in which he is considered as a black box. If real neuron actions will become observable, the "box" will become transparent and real uploading will be possible.
There are several steps in the practical part:
- The first step includes all the methods of fixing information while the person of interest is alive.
- The second step is about preservation of the information.
- The third step is about what should be done to improve and promote the process.
- The final fourth step is about the reconstruction of the individual, which will be performed by AI after his death. In fact it may happen soon, may be in next 20-50 years.
There are several unknowns in DI, including the identity problem, the size and type of information required to create an exact model of the person, and the required power of future AI to operate the process. These and other problems are listed in the box on the right corner of the map.
The pdf of the map is here, and jpg is below.
Previous posts with maps:
Doomsday Argument Map
AGI Safety Solutions Map
A map: AI failures modes and levels
A Roadmap: How to Survive the End of the Universe
A map: Typology of human extinction risks
Roadmap: Plan of Action to Prevent Human Extinction Risks
Immortality Roadmap