3 min read28th Jul 201527 comments

20

Added: Direct link on pdf: http://immortality-roadmap.com/IMMORTEN.pdf

 

A lot of people value indefinite life extension, but most have their own preferred method of achieving it. The goal of this map is to present all known ways of radical life extension in an orderly and useful way.

A rational person could choose to implement all of these plans or to concentrate only on one of them, depending on his available resources, age and situation. Such actions may be personal or social; both are necessary.

The roadmap consists of several plans; each of them acts as insurance in the case of failure of the previous plan. (The roadmap has a similar structure to the "Plan of action to prevent human extinction risks".) The first two plans contain two rows, one of which represents personal actions or medical procedures, and the other represents any collective activity required.

Plan A. The most obvious way to reach immortality is to survive until the creation of Friendly AI; in that case if you are young enough and optimistic enough, you can simply do nothing – or just fund MIRI. However, if you are older, you have to jump from one method of life extension to the next as they become available. So plan A is a relay race of life extension methods, until the problem of death is solved.

This plan includes actions to defeat aging, to grow and replace diseased organs with new bioengineered ones, to get a nanotech body and in the end to be scanned into a computer. It is an optimized sequence of events, and depends on two things – your personal actions (such as regular medical checkups), and collective actions such as civil activism and scientific research funding.

Plan B. However, if Plan A fails, i.e. if you die before the creation of superintelligence, there is Plan B, which is cryonics. Some simple steps can be taken now, such as calling your nearest cryocompany about a contract.

Plan C. Unfortunately, cryonics could also fail, and in that case Plan C is invoked. Of course it is much worse – less reliable and less proven. Plan C is so-called digital immortality, where one could be returned to life based on existing recorded information about that person. It is not a particularly good plan, because we are not sure how to solve the identity problem which will arise, and we don’t know if the collected amount of information would be enough. But it is still better than nothing.

Plan D. Lastly, if Plan C fails, we have Plan D. It is not a plan in fact, it is just hope or a bet that immortality already exists somehow: perhaps there is quantum immortality, or perhaps future AI will bring us back to life.

The first three plans demand particular actions now: we need to prepare for all of them simultaneously. All of the plans will lead to the same result: our minds will be uploaded into a computer with help of highly developed AI.

The plans could also help each other. Digital immortality data may help to fill any gaps in the memory of a cryopreserved person. Also cryonics is raising chances that quantum immortality will result in something useful: you have more chance of being cryopreserved and successfully revived than living naturally until you are 120 years old.

After you have become immortal with the help of Friendly AI you might exist until the end of the Universe or even beyond – see my map “How to prevent the end of the Universe”.

A map of currently available methods of life extension is a sub-map of this one and will published later.

The map was made in collaboration with Maria Konovalenko and Michael Batin and its earlier version was presented in August 2014 in Aubrey de Grey’s conference Rejuvenation Biotechnology.

Pdf of the map is here

Previous posts:

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

 

 

 

 

 

 

(scroll down to see the map)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(scroll down to see the map)

 

 

 

 

 

 

 

 

 

 

 

 

(scroll down to see the map)

 

 

 

 

 

 

 

 

 

 

New Comment
27 comments, sorted by Click to highlight new comments since: Today at 8:34 PM

Overall, the strategy and organization are good. Where it goes terribly wrong is the completely unfounded optimism that SENS will start to pan out in 15 years and that nanobots will arrive in 25 years. For short-term technology, consensus from the experts is usually very accurate. The experts are nearly unanimous that both of those technologies are theoretically possible, but are so far in the future as to be irrelevant. Aubrey de Grey is irrational on this issue. I don't want to be too hard on him, as he is one of the very few in the world to accept cryonics, but there is simply no basis for the false optimism.

I explain it better here: http://www.oregoncryo.com/revivalTechnology.html I put nanobots at 180 years out to emphasize how much harder that technology is compared to other forms of nanotechnology.

Timeline is rather arbitrary and yes it is optimistic. It may shift right depending of the pace of technological progress. I also agree with you view on Aubrey. In my opinion nanobots may appear much quicker if they will be built by augmentation of biological cells. If one combine E.coli and already exiting DNA origami tech, he could get universal mechanism able to self-replication.

I'd like to retract my comments to some degree. I think the problem is definitions.

"SENS: Regular correction of accumulated damage." You can't just put that on a timeline without qualifications. There are 7 complex problems to be solved in SENS. If each one of them takes 10 years to solve, you're talking about SENS breakthroughs spanning 70 years. The only claim you can make for 15 years out is maybe one small breakthrough that partially addresses one of the seven problems. It would be nothing close to your characterization and wouldn't help us very much.

"Non self-replicating nanobots" I want to make it clear that you're talking about crude first generation nanobots, while I was talking about fully mature nanobots capable of cryonics repair. The two technologies are lightyears apart in complexity, capability, and timeframe. There could easily be 100 or more years of development separating the two. But your wording still seems to indicate intelligent swimmers with manipulators, which is more of an intermediate technology rather than first generation.

It's worth mentioning that a lot of the kinds of technology in path A could also lead to immortality (or at least, to extreme longevity, "immortality" is kind of a loaded word) if strong AI either doesn't happen at all, or happens but for some reason doesn't help as much as we think, or if we decide to not do it at all, or if it takes much longer then we expect to happen. I don't know how likely any of those possibilities are, but I don't think they're all zero.

I think that without uploading or creating copies (and without AI) any life extension method would give 1000 years life expectancy maximum. It is very long but nowhere near immortality.

Uploading is almost like creating AI, so if we will have EMs, we will come to AI very soon after it. In the earlier versions of the map plan A was named "Victory on aging", and after it was renamed in "To survive until Immortality", which is oxymoron, so I think that "survive until FAI" is the best name and idea.

Uploading is almost like creating AI, so if we will have EMs

It should be "ems" as it is short for emulation, not an acronym. Sorry for the nitpick, I've seen this a lot and it bothers me for some reason.

I think that without uploading or creating copies (and without AI) any life extension method would give 1000 years life expectancy maximum

That was Aubry de Grey's estimate of how long people would live if we cured aging, but that assumes the same rate of deaths from accidents, suicides, ect as we have today. But with advanced cyborg technology, medical technology, biotechnology, and/or nanotechnology, as well as other technological improvements that reduce the risk of accidental death (self driving cars, for example), that could be a lot longer.

Edit: To be clear, I agree with you that if uploading/"making copies" proves possible, that that could theoretically grant a much "safer" kind of longevity that could last much longer on average. I'm just saying that even without that, 1000 years is just a starting point towards what's possible, and we likely be able to do significantly better then that.

[-][anonymous]9y30

Great map! It's very weird to me that plan A is survive until friendly AI. It assumes that we'll solve friendly AI before we solve Aging. Seems like an unwarranted assumption, especially if we can start seeing breakthroughs in scientific research through machine learning, sans general AI. Im also wondering where EMs are. Keep up the great work

No, It doesn't assumes that we solve aging before AI. In fact we need to solve aging to survive until the moment than FAI will be created, which I put optimistically in period 2050-2100. I think that EM is a type of AI. Brain scanning in the end of plan A will result in EM.

In case you didn't know, storing writing as images like it is in your mind map is bad for accessibility. Those who aren't visual, for example search engine indexing bots and blind humans, have difficulty reading such writing.

http://immortality-roadmap.com/IMMORTEN.pdf

The pdf is here. But I will upload it in google docs also which hopely will improve searchability.

I am storing it as a pdf, which is linked on the top of the post, and added image screen shot just to show what is it, but I know that there is a problem how to present large plots in a post if the original is pdf.

I like these maps. Keep up the good work!

The pdf is fine, but the image version of this map is unreadable. It might be worth making a higher resolution version of it.

Ok, I will try to upload high resolution version, thanks for your support. As I get many unexplained downvotes, any positive comment helps me to continue the work on the maps.

I have a question about the digital reconstruction/revival/uploading/etc. How do you know it's truly you? Do you wake up and just see you're now in a computer?

I might be missing something, but the process in my head (heh) is that the brain is being copied, analyzed and then booted on the computer. Then there's two copies of the person, and while the biological brain will eventually die, the brain ex machina will still be living.

However, the brain ex machina, even though it's an identical copy, is not the same person. Therefore, it's more like personality-immortality, or brain-data-immortality. The person's conscious experience will be branched off until the brain dies.

I thought immortality implies that you actually stay conscious, not that you die for a hundred percent but are preserved.

The reconstruction plan is divided in two tasks: creating exact model of person and solving identity problem. I assume here that in the future identity problem will be solved. In this map I just do not touch it. But I have separate map in progress (and 100 pages text) where I try to address this problem.

Basically, solving identity problem depends of how one solve problem of consciousness and nature of reality. If one assume that mind is only information, than where is no problems with copies. Any human have many future copies, like me-tomorrow me-day-after-tomorrow.

If you have a model of mind, where identity depends of something which is not information, you must invest in preserving exactly it, but not the information. This thing may be named "essence of identity", and most popular such essence is "continuity of consciousness" (But also it may be qualia, soul etc).

Anyway I hope that even if the essence of identity is real, we will built separate mechanisms to transfer it from the brain to the computer (like gradual uploading or preserving small part of the brain with active electric process). But as the problem of identity is not solved yet, the plan C is poor and should be done only after plans A and B fails.

If my brain is dying it is better to have my copy in a computer than don't have anything. Using indexical trick as identity transfer mechanism may also help in this case: if I don't know am I in computer or not.

[-][anonymous]9y00

Where you draw the line on these sorts of things is personal preference. The "copy" can't tell he's a copy.

You suggested reading Longecity. However, it seems that most articles on Longecity are only relevant to whoever posted it, for example by asking about what to do in a very specific situation, or aren't relevant at all to increasing the chance of becoming immortal. Knowing this, how do you recommend reading Longecity, if at all?

Hi! I have been reading it for years to find useful leads and also used search if I have any specific topic. The posts there are not equal, some has many links on scientific articles, like Lostfalco posts, some just personal experiences.

Thanks for the post. The bottom of the mind map references the book Immortality by Alexey Turchin, but an Internet search failed to reveal any links to or discussing it. Do you know where it can be found?

The poster is the person who wrote the book.

Yes, that's why I'm asking here.

Hi!

The book is in draft version and in Russian, but some of its ideas are presented in my roadmaps about immortality which are in fact chapters from the book.

The text of the book if here: https://drive.google.com/open?id=1XZ548GMiV2dmdIY_xDyOmwv-pi2SouQZmx8-sOLknhc

The links on the other roadmaps are here: http://immortality-roadmap.com/sample-page/

If a version is written in English, I'll probably be willing to review and proofread it. I'm a decent writer in English, and I know a fairly large amount about immortality. I wrote Immortality: A Practical Guide using a different Less Wrong account, in case you're interested.

Hi It may take a couple of years to finish it and translation may need substantial adaptation. that is partically because I am working now on a series of article on global risks prevention. But some chapters and parts of it may be translated earlier. And I would appreciate you help. If you have facebook account, you may found me there as Alexey Turchin.

I am now working on a chapter "AI in life extension", which is hot topic, and may be translated soon. Proofreading is always helpful as I am not native English speaker. I may ask you for help in it.

Okay, when you would like me to help, email me at 0xakaria@gmail.com.

Thanks for the response. Do you know if the book will be made available in English, and if so, approximately when?