This may sound pessimistic, but the value of brain preservation depends also on your views about the long term future. If you think that there is a non-negligible chance that the future is ruled by a non-aligned AI or that it will be easily possible to create suffering copies of you, then it would be better to erase the information that is necessary to reconstruct you after your biological death.
Good point. This is true for those who believe this, but it applies to any form of medicine or life extension, right? Not just brain preservation. So for someone who holds this view, theoretically it might also apply to the antimalarial medication case as well?
The difference is that if you are biologically dead, there is nothing you can do to prevent a malevolant actor to upload your mind. If you are terminally ill and are pessimistic about the future, you can at least choose cremation.
I am not saying that there should be no funding for brain preservation, but personally I am not very enthusiastic since there is the danger that we will not solve the alignment problem.
I'm not sure I understand the scenario you are discussing. In your scenario, it sounds like you're positing a malevolent non-aligned AI that would forcibly upload and create suffering copies of people. Obviously, this is an almost unfathomably horrific hypothetical scenario which we should all try to prevent if we can. One thing I don't understand about the scenario you are describing is why this forcible uploading would only happen to people who are legally dead and preserved at the time, but not anyone living at the time.
To answer your question I have to describe some scenarios how a non-aligned AI would act. This is slightly cringe since we do not know what an unaligned AI would do and this sounds very sci-fi like. In the case of something like a robot uprising or a nuclear war started by an AI many people would die under circumstances such that uploading is impossible, but brain banks could be still intact. If an unaligned AI really has the aim to upload and torture everyone, there will probably be better ways. [Insert something with nanobots here.]
In my personal, very subjective opinion there is a 10% chance of extinction by AI and a 1% chance for s-risks or something like Roko's basilisk. You may have different subjective probabilities and even if we agree on the possibilities, it depends very much on your preferred ethical theory what to do.
I'm not disagreeing with you that there is a possibility, however small, of s-risk scenarios. I agree with this point of yours, although I'm thinking of things more like superhuman persuasion, deception, pinning human tribes against one another, etc., rather than nanobots necessarily:
If an unaligned AI really has the aim to upload and torture everyone, there will probably be better ways. [Insert something with nanobots here.]
People often bring up this in the context of brain preservation. But it just seems to me that this possibility is mostly a generalized argument against life extension, medicine, pronatalism, etc in general.
On the one hand, I am more concerned with learning AI safety than organizing my own cryoperservation or chemical brain preservation. This is mostly because I think it is more important to avoid the worst possible futures before planning to be alive in the future. On the other hand, I think life extension should receive much more funding (as long as you do not use AI with dangerous capabilities), I am against anti-natalism and I think that brain preservation should be available for those who wish. There is a certain tension between both statements. I do not know how to resolve that tension, but thank you for pointing out the problem with my worldview.
Executive summary: Brain preservation via chemical fixation and fluid storage is a speculative but potentially cost-effective intervention for saving lives in the long-term, depending on key assumptions about the probability of successful future revival.
Key points:
This comment was auto-generated by the EA Forum Team. Feel free to point out issues with this summary by replying to the comment, and contact us if you have feedback.
Summary: This post argues that brain preservation via fluid preservation could potentially be a cost-effective method to save lives, meriting more consideration as an EA cause. I review the current technology, estimate its cost-effectiveness under various assumptions, and explore key considerations regarding its feasibility and value.
Previous EA discussions include: here and here
Disclosure: I am employed as a research scientist at Oregon Brain Preservation and I am a volunteer director of Apex Neuroscience.
Intro
Brain preservation is the process of stabilizing and protecting the physical structure of the brain after legal death, with the goal of allowing the individual to be "revived" in the future if the necessary technology, such as molecular nanotechnology or whole brain emulation, ever becomes available.
There are three main current approaches to brain preservation: pure cryopreservation, aldehyde-stabilized cryopreservation, and chemical fixation followed by fluid preservation. All three methods aim to preserve the intricate synaptic connections and biomolecular details across the brain that are thought to encode an individual's memories, personality, and personal identity.
Pure cryopreservation is the traditional method used in cryonics, where the brain is perfused with cryoprotectants and cooled to very low temperatures, typically using liquid nitrogen. Aldehyde-stabilized cryopreservation, on the other hand, involves an initial step of chemical fixation, followed by the distribution of cryoprotectants, followed by very low temperature cooldown. This additional fixation step is thought to improve the structural preservation of the brain compared to pure cryopreservation.
Finally, chemical fixation followed by fluid preservation involves fixing the brain tissue and then storing it long-term in a bath of preservative chemicals at room temperature or refrigerator temperature. This is the most affordable option for long-term storage so it is what we will discuss here.
Technical assumptions for brain preservation via fluid preservation
1: The brain is "all you need", i.e. it contains all the necessary information for an individual's memories, personality, and identity. If the brain can be preserved, the rest of the body could potentially be replaced or regenerated using advanced technologies such as 3D bioprinting, robotic prosthetics, or emulation. This simplifies the preservation process and decreases costs substantially.
2: Chemical fixation can preserve neural structures and biomolecules throughout the brain in close to their original locations. This is likely what is required for the information in memories and other valued cognitive functions.
3: After chemical fixation, the brain can be stored in a preservative solution at room temperature or refrigerated conditions. This fluid preservation is assumed to provide a stable environment that prevents significant deterioration of the fixed neural structures over extended periods. The reliability and affordability of this storage method are key. However, the long-term effectiveness of fluid preservation in maintaining the integrity of neural structures is an area of uncertainty.
4: The necessary technologies for reviving a person from their preserved brain in a humane way are ever developed in the future.
5: It is assumed, as with many technologies, that initial high costs for the revival method could rapidly decrease as the techniques are refined and automated. Therefore, a relatively small philanthropic investment could potentially cover the revival costs.
Each of these assumptions has its own challenges and uncertainties that need to be carefully considered when evaluating the feasibility and potential of brain preservation via fluid preservation.
Cost of brain preservation via storage in liquid
The average cost of brain isolation and immersion fixation in the United States is estimated by one source to be around $1000 [1]. This represents the upfront cost of preserving a brain, including both labor and supplies. The most expensive form of equipment is probably the bone saw, which can cost up to a few thousand dollars, but can be used many times. Formaldehyde solutions themselves are inexpensive and not a major cost contributor. The labor itself is the most expensive part of the cost, which would likely be lower in lower-income countries, making brain preservation potentially more cost-effective there.
The ongoing storage costs for a preserved brain appear to be quite low. Brain banks routinely store brains in liquid in plastic containers on shelves for extended periods at minimal expense. For example, the University of Geneva brain bank has more than 10,000 brains stored in liquid, which have been stored for many decades. An estimate for long-term storage might be $300 per person whose brain is preserved.
Finally, there is the cost consideration of reviving a person whose brain has been preserved. Currently, this is simply not possible, and it may never be. It is likely that when this technology is first invented, it will be very expensive, but it will likely be automated relatively quickly, as it is likely to involve the heavy application of machines. Consider, for example, how difficult -- or even inconceivable -- whole genome sequencing would have seemed 100 years ago, but now it is possible for $200, and the price is continuing to drop. In the long term, perhaps $200 of present day dollars, invested with compound interest, would be sufficient for a revival procedure, again assuming it is ever possible.
Subjective probability of revival in the future
This is obviously the most challenging factor to estimate here and is a key area for further research. Opinions vary widely on the likelihood that a preserved brain could be successfully revived in the future with memories and identity intact, and the extent to which it depends on the method used for brain preservation. Some view it as almost inevitable given the trajectory of technological progress, especially in AI, while others see it as highly improbable or even impossible. Given the current state of scientific understanding, any estimates of this probability are necessarily highly subjective and uncertain. As a placeholder for the sake of analysis, we consider a wide range of subjective probabilities from 0.01% to 75%, but these should be viewed as highly tentative and open to revision.
Cost of saving a life through other means
To provide a point of comparison, we can look at the cost-effectiveness of other life-saving interventions. GiveWell, a respected charity evaluator, estimates that it costs around $3000 to save a life by distributing antimalarial medicines in areas where malaria is highly seasonal.
Comparative cost-effectiveness
Putting these pieces together, we can estimate the cost-effectiveness of brain preservation via fluid preservation in terms of "lives saved" under different scenarios:
At a 75% subjective probability of successful revival, the cost per life saved would be around $2000 (= ($1000 + $300 + $200) * 0.75).
At a 50% subjective probability of successful revival, the cost per life saved would be around $3000, putting it on par with distributing antimalarial medicines.
At a 10% probability, the cost per life saved would be around $15,000.
At a 1% probability, the cost per life saved would be around $150,000.
At a 0.1% probability, the cost per life saved would be around $1,500,000.
At a 0.01% probability, the cost per life saved would be around $15,000,000.
These numbers are, of course, highly speculative and uncertain. The purpose here is not to make strong claims about the actual cost-effectiveness of brain preservation, but rather to illustrate how it could potentially compare to other interventions under different sets of assumptions. I discuss some of the open questions that make this so hard to estimate here.
It's also worth noting that this analysis does not account for the other potential positive or negative externalities of the practice on society. For example, the potential for brain preservation could help give people more of a stake in the future. Also, preserved individuals could contribute their wisdom and historical knowledge to future societies. On the other hand, it could exacerbate inequality if brain preservation is only accessible to the wealthy.
As the world gets richer and more resources flow to global health and poverty causes, Givewell estimates that interventions aimed at saving lives in the near-term may become more expensive. In that scenario, brain preservation could become relatively more appealing from a cost-effectiveness perspective.
In the long run, regulations could emerge around brain preservation specifically, which could be really beneficial for quality, but also make it more expensive. But that's hopefully kind of a win scenario in terms of societal focus on this topic, and anyway seems quite unlikely anytime soon, so not worthy of much consideration.
Of course, these estimates are highly uncertain and speculative. We should be cautious about putting too much weight on them. Nevertheless, they suggest that brain preservation is at least within the realm of being a competitive altruistic intervention under some sets of assumptions. More research on the key parameters -- especially the probability of successful revival -- could help to refine these estimates.
Four perspectives on the altruistic value of brain preservation
Given the uncertainties and potential societal impacts, people may have different views on the altruistic value of brain preservation. Here are four perspectives:
1: Preserve now, research is secondary
If one believes the subjective probability of success is already high enough, the altruistic priority should be to focus resources for preserving people who currently want this option. While more research could be beneficial, the primary goal should be to make the best use of existing preservation technology. Every delay means more people may be irreversibly lost.
To keep costs as low as possible, preservation would be best performed by people working in a local area. The cost of deploying team members to remote locations, obtaining the necessary space and equipment, and transporting bodies quickly adds up.
2: Research first to inform decision-making, maybe preserve later
For those who see the probability of success as highly uncertain, the value of information from further research may be the most important parameter. Investing in studies to better understand the feasibility and implications of brain preservation could help guide decisions about whether and how much to prioritize it. It would be premature to spend significant resources on preservation until we have a clearer picture. Such research could also lead to methodological advances that improve the subjective probability of success.
3: Don't pursue, because success is too unlikely
If someone estimates the odds of eventual revival to be extremely low, even with further research, then brain preservation would not be a worthy use of limited altruistic resources. Those funds would be better spent on more tangible ways to improve lives. Wishful thinking should not divert us from clear opportunities to do good.
4: Avoid, even if feasible
Some may hold that successfully reviving preserved brains would be a net negative for society, regardless of the probability it could work. Concerns could include overpopulation, aiding in totalitarian lock-in, cultural ossification, or the hubris of defying natural death. From this view, brain preservation should not be pursued on principle, irrespective of its technical feasibility.
Beyond these positions, there are additional considerations. For example, some argue that "lives saved" is an inadequate metric and that there is a notion of a "complete life" -- perhaps around 60 years. On this view, extending life beyond that point carries diminishing moral value compared to saving younger lives.
Considerations for funding
When I was doing research at Mount Sinai, I discussed this topic with two senior researchers who each had a ton of experience in grant writing and thought that the idea of brain preservation was good in general (or at least they told me so, after learning that I was highly interested in the field). They both told me that there was effectively a zero percent chance that a government agency would fund a brain preservation research program, at least anytime soon. Brain preservation is considered too speculative and unconventional to fit within the funding priorities of these agencies. This means that funding would have to come from a non-governmental source.
Brain preservation is often seen as a fringe or controversial idea, which may contribute to its neglectedness in terms of research funding. Offering brain preservation for free could help mitigate skepticism and perceived scamminess. However, a free option would require philanthropic funding to cover labor and equipment costs. Therefore, philanthropic funding could be crucial for advancing the field.
For those interested in supporting the development of brain preservation, donating to organizations in the field could potentially be a high-impact way to contribute, again depending upon one's perspective on the altruistic value of brain preservation.
Non-monetary support for brain preservation
A large number of the barriers to successful brain preservation are not necessarily amenable to scientific research, but rather require policy changes. The reason I point this out is that even if individual EAs decide that brain preservation is not the best use of their donation dollars, there are still important ways they could potentially support the development of this technology through advocacy and social influence.
For example, people are not able to opt out of the medicolegal delay required after a legal death that is not classified as occurring due to natural causes. This can go on for many hours, during which the brain tissue will be degrading, before the preservation team is able to start the procedure. In the worst case scenario, this leads to an involuntary autopsy, which obviously has the clear potential to dramatically worsen brain preservation quality.
People in most jurisdictions are also not allowed to time the procedure when they want to. The only reason it is possible in some areas is because medical aid in dying is already legal there, and people can dovetail off of this. But this is actually not ideal either for a variety of technical reasons. Theoretically, the highest quality brain preservation would be more integrated with a euthanasia procedure, where this is legal and available to an individual [2].
Probably the most useful thing would be a cultural shift in the way that brain preservation is perceived, including via outreach in scholarly communities. There is a petition on change.org that discusses more of the ways that brain preservation is being held back by current societal laws and norms.
Key questions to explore
To better assess the potential of brain preservation as an EA cause area, I highlight a few of the key questions and areas for further research.
1: Neural correlates of identity. What are the exact neural structures and patterns that encode an individual's memories, personality, and sense of self? Answering these questions would help clarify the key information that brain preservation needs to capture to enable future revival of a person. There doesn't seem to be an overwhelming consensus on this yet within the field of the neuroscience of memory.
2: Preservation quality. To what extent do current brain preservation methods, particularly chemical fixation and room-temperature storage in liquid, preserve the neural correlates of identity identified above? Systematic studies comparing preserved brain tissue with different methods at multiple scales (from molecular to anatomical) could help measure preservation quality.
3: Revival methods. What, if any, are the most plausible pathways for reviving a person whose brain has been preserved? Analyzing the feasibility, requirements, and limitations of different revival scenarios could inform R&D priorities and timelines.
4: Civilizational factors. How might long-term sociological, economic, and geopolitical trends, such as the risk of civilizational collapse or the development of transformative AI, affect the likelihood that a preserved brain could be revived in the future?
5: Ideal vs. practical cases. How does the probability of successful revival vary between ideal conditions -- i.e. immediate high-quality preservation -- and more realistic scenarios involving delays, lower-quality preservation, or procedural mistakes by the preservation team? One possibility here is that only brain preservation under certain conditions, such as cases with a relatively minimal delay after cardiac arrest, might meet a particular set of standards for effectiveness as a use of altruistic funds.
6: Ethical and philosophical considerations. Beyond technical feasibility, what are the key ethical and philosophical considerations around brain preservation and future revival? How might these perspectives evolve over time, as societal priorities change?
Conclusion
Brain preservation is undoubtedly a controversial and speculative idea. Many may dismiss it outright as infeasible or not worth considering.
As a community committed to rational analysis, open-mindedness, and the pursuit of truth, EAs have a unique opportunity to lead the conversation on unconventional ideas like brain preservation. This isn't to say that brain preservation will necessarily prove to be a worthwhile or feasible intervention. But if the EA community doesn't even consider it, there is a risk of missing out on potentially high-impact opportunities, which may eventually look unwise with the benefit of hindsight.
It seems to me that if there's even a modest chance that brain preservation could be a cost-effective way to save lives then it is worthy of further examination and debate. At the very least it seems that there should be detailed arguments available for why brain preservation has a low subjective probability of success, which is not currently the case.
I'm particularly curious to hear what others on the EA forum think about this topic and why they disagree with me, if they do.
This assumes that immersion fixation would be sufficient for preserving neural structures even in inner areas of the brain. As far as I can tell, this is an open research question. If not, perfusion of the fixative chemicals would be required, the upfront cost would be more expensive, and likely not at parity with other contemporary options, even assuming a very high subjective probability of revival in the future.
This option, if available, would raise important ethical questions. It would be crucial that the decision to pursue euthanasia be made based on a careful consideration of one's overall quality of life and personal values, strictly independent of brain preservation providers. Euthanasia and brain preservation are both deeply personal decisions that require thoughtful deliberation and should not be unduly influenced by external pressures or the promise of a particular outcome.
Thanks Andy for writing this up! I have two comments:
1. For those interested in a longer yet accessible treatment of this topic written for a general audience, I've got a book coming out on this in November - The Future Loves You: How and Why We Should Abolish Death. It covers much of the medicine, neuroscience, philosophy, and economics that Andy already touched on in this post, plus more besides.
2. One thing Andy didn't touch on and that I don't cover in my book is that the ethical imperative of brain preservation seems to be very different based on whether one holds a 'total hedonic' view or a 'person-affecting' view of utilitarianism. In a total view, the particular individuals having the happy experiences don't matter, just that there are happy experiences being had. In the person-affecting view, one cares more that particular people have happy experiences, and it would be bad if everyone on Earth suddenly ceased to exist and were switched out for completely new people of equal or greater total happiness. I think "normal" people tend towards a person-affecting view, while effective altruists typically tend towards a total view. I'd be curious to see a formal analysis of how brain preservation fits into these views though.