Though cryonics has been practiced for forty years, its techniques have improved only slowly; its few customers can only induce a tiny research effort. The much larger brain research community, in contrast, has been rapidly improving their ways to do fast cheap detailed 3D brain scans, and to prepare samples for such scans. You see, brain researchers need ways to stop brain samples from changing, and to be strong against scanning disruptions, just so they can study brain samples at their leisure.

These brain research techniques have now reached two key milestones:

  1. They’ve found new ways to “fix” brain samples by filling them with plastic, ways that seem impressively reliable, resilient, and long lasting, and which work on large brain volumes (e.g., here). Such plastination techniques seem close to being able to save enough info in entire brains for centuries, without needing continual care. Just dumping a plastic brain in a box in a closet might work fine.
  2. Today, for a few tens of thousands of dollars, less than the price charged for one cryonics customer, it is feasible to have independent lab(s) take random samples from whole mouse or human brains preserved via either cryonics or plastination, and do high (5nm) resolution 3D scans to map out thousands of neighboring cells, their connections, and connection strengths, to test if either of these approaches clearly preserve such key brain info.

An anonymous donor has actually funded a $100K Brain Preservation Prize, paid to the first team(s) to pass this test on a human brain, with a quarter of the prize going to those that first pass the test on a mouse brain. Cryonics and plastination teams have already submitted whole mouse brains to be tested. The only hitch is that the prize organization needs money (~25-50K$) to actually do the tests!

Comments?  If superior brain preservation can be demonstrated under a 5nm-resolution 3D scan, plastination wins over vitrification hands-down.  Is Robin missing anything here, or is this indeed as important as he says?

New to LessWrong?

New Comment
47 comments, sorted by Click to highlight new comments since: Today at 10:44 AM

Is Robin missing anything here, or is this indeed as important as he says?

I think it's as important as he says; I was very pleased to see that he has gone through with it, and I've donated $50 to the Evaluation Fund.

It loses any data that is not structural in the neurons' physical shape - whose importance is not currently known. We can presume that electrical signals can be rebooted, but can chemical ones? Will the brain fail as badly as a drunkard or someone who drinks twenty espressos, if shorn of its chemical context?

This is particularly plausible because the brain is full of low-level feedback loops controlling endocrine stuff - I could fully expect them to go completely bugfuck if their sensor inputs suddenly read "0.0".

To give an example here: "gonadotropin-releasing hormone analogue" drugs are used to block secretion of sex hormones. They were originally designed to increase them. GNRH is the "on switch" signal, the drugs mimic it. And indeed they do initially increase hormones - then the brain's regulatory feedback slams on the brakes, all the way to zero. Nobody knew that particular feedback was there before they poked it with a drug.

This tech may make testing the above much easier though.

Freeze, slice, stain, and microscope can check chemicals in the way this cannot.

So... set a reasonable endocrine state, not 0.

There are few enough degrees of freedom in the endocrine system that having it arbitrarily reset to some normal level... won't alter your identity, it'll just alter your mood.

Especially since this is going to have to be emulated anyway.

That works fine if it's just the endocrine system whose information you lose. It works less well if there are a hundred systems, each with several degrees of freedom, whose joint state needs to be in some relatively narrow region in order to preserve one's identity.

Yeah, but my point being, you don't know all those feedbacks. There are probably scads of them. And realistically the only way to find out would be to boot up a great many nonhuman animals first, and watch them bug out in informative ways. Which is likely to be cruel work, and not fun at all.

Also "only alter your mood"? Well, only as much as being hypoglycemic, or hypoxic, or panicking for breath, or ravenous, or various other hormone or feedback linked things alter your mood, especially with them all firing at OHSHITGONNADIE levels all at once - ie, it would be instantly and horribly incapacitating.

It wouldn't be nice at all, but it wouldn't be death, and it would be fixable even with current psych meds.

No, it would not. Those meds rebalance an unbalanced system, they don't rebuild a completely undercut system. And you are assuming that simulating the chemical effect of the meds would be easy - lest you forget, we've thrown away the chemicals, that's the problem.

Your claim seems important if true. My personal tendency is to preserve data where I don't understand what's going on, and my impression is that lots of data is stored in the biochemistry itself... precisely which receptors exist, how deeply they're embedded in membranes, what kinds of fatty acids a membrane is made of, and so on. Dendrites and axons engage in chemotaxis where they follow genetically programmed ion and/or protein gradients. If you wipe out the chemical gradient information then how do you know what sorts of ways that the dendrites should regrow in the weeks and months post-resuscitation?

A divergence of opinions here could be due to different assumptions about what kind of knowledge will be available in the context of resuscitation. If your resuscitators have vast and deeply tested knowledge of the operation of many kinds of brains, then they might be able to mock up plausible values for all the chemistry and produce someone "similar enough" (and its certainly better than nothing but a bunch of letters and photos and Madeleines). If they're doing this for the first time in a sort of "1950's plucky engineer mode", then I would naively suspect that plastination pushes someone closer to information theoretic death than even membrane-damaging no-vitrification cryonics of the sort that was standard operating procedure in the mid-1970s.

That's pretty much my take on it, though with different details.

Is the idea to upload the record on to a generic brain? Do the details of DNA and its methylation matter to identity?

If you wipe out the chemical gradient information then how do you know what sorts of ways that the dendrites should regrow in the weeks and months post-resuscitation?

If I wake up and I feel like myself on a second to second basis, I will not be upset if my path through mind space is drastically altered on a time scale of weeks and months, so long as it doesn't lead me to insanity. Hell, I hope I'll be able to drastically change my mind on that time scale anyway once I'm uploaded.

This will end up being important if the details of these systems are important for encoding the differences between humans rather than just important for having a properly functioning brain. If it's just important for proper functioning, we can just figure it out once and then assume that the brain you're reconstructing has these systems in good working order. If it's important for encoding the differences between humans then we'd have to preserve these systems to preserve "you".

I'm reminded of Charles Stross on space colonization, where he talks about how it's a bit too late to realize you forgot the (insert essential mineral here) supplement when your interstellar generation ship starts coming down with the purple polkadot scurvy at 0.001c and boosting. There's a reason we can't reliably provision a generation ship, and it's that we have never yet tried to completely and permanently sever ourselves from Earth's ecology and biosphere. We may think we've got it all covered, but if there's a leak in the cycles somewhere, or something missing we never knew was important, our intrepid astronauts are going to be in for a hard time, either immediately or generations later.

This by analogy strikes me as a general problem with uploading, but a specific problem with anything that throws away a lot of "body biosphere". There will be an initial shakedown period, mostly on animal models, where we learn the obvious breakages (some of which are likely to only show up in human uploads because they create subtler kinds of mental illness). But it's going to be hard to be sure we have eliminated all the deficiencies and closed all the feedback loops. It will just plain take time, and a lot of unpleasantness and health scares.

There's a reason we can't reliably provision a generation ship, and it's that we have never yet tried to completely and permanently sever ourselves from Earth's ecology and biosphere. We may think we've got it all covered, but if there's a leak in the cycles somewhere, or something missing we never knew was important, our intrepid astronauts are going to be in for a hard time, either immediately or generations later.

So redo Biosphere 2, for longer. Even the first time it was done, it worked remarkably well! They did make some mistakes but recovered, and came out healthier than they went in. That suggests a multi-generation-capable version is not as far off as one might pessmistically conclude. The most valuable information is always the first information - if a problem doesn't appear quickly, then it probably isn't that important...

if a problem doesn't appear quickly, then it probably isn't that important...

I agree completely, especially about how close we probably are to a successful Biosphere, but just to throw out an example where this is wrong: vitamin B-12 deficiency usually takes a decade to demonstrate symptoms, and is fatal.

This seems important enough for Main to be honest.

I agree. This is pretty exciting.

A few comments:

1) Sign up for cryonics now. Do not delay because you think "Plastination might be better someday in the future".

  • No one is offering plastination now, and you can die now.
  • It's not clear when, if ever, anyone is going to offer plastination.
  • It's not clear that plastination, if actually offered, will actually be better than cryonics.

2) With chemical preservation, good vascular perfusion is critical. Brain tissue which is not perfused is lost.

3) With cryopreservation, good vascular perfusion results in excellent preservation by vitrification. Brain tissue which is not perfused is still preserved.

4) Most of the costs for neuropreservation are from the up-front logistical and surgical costs. If you want minimal ischemic time and good vascular perfusion for chemical preservation you're still going to have to pay most if not all of those costs.

Wins for what? I don't think plastination is an option for human preservation today. When it becomes an option, it probably wins.

The problem with plastination is scaling up the volume that can be done at once. This is a matter of pumping fluids around. Tiny chunks of mouse brain that were plastinated 50 years ago have readable synapses today. The experiment is whether new methods of applying chemicals to whole mouse brains work as well as first cutting up the brain; and whether cutting after plastination preserves enough information.

After scaling up plastination, it has the remaining downside that it displaces lots of chemicals. RH asks to preserve "two dozen chemical densities," which it probably fails at. Also, lots of sub-synapse detail (eg, type and placement of ion channels) is probably lost.

Also, lots of sub-synapse detail (eg, type and placement of ion channels) is probably lost.

Are you sure? If I recall correctly (it's been a long time), a typical ion channel consists of 10^4-10^6 atoms and is something like 1-10nm in size, which is about the same as the current 3D scan resolution, and it is reasonably easy to infer the type and abundance of channels from their location. In any case, even the current resolution is almost enough to identify channel location, and only a small increase in resolution would be enough to identify channel types.

No, I'm not sure. But I'm not talking about the limits of current read-out techniques, which will of course improve. I'm talking about the destruction of information by the process of plastination: chemical change of the membrane and replacement of almost everything else. Currently what is read is the shape of the membrane. The location, if not type, of an ion channel is probably readable from distortions of the membrame. They might be left intact, trapped by the membrane, readable by future techniques. But I wouldn't be terribly surprised if they are completely wiped out.

Most likely, we don't need to know, but can guess based on other details of the synapse.

(This comment is largely repeating something from my blog)

I would suggest storing, along with the brain, a representative "snapshot" of the working brain, possibly an EEG under standardized conditions.

In the cryonics model, storing your EEG's didn't make much sense. When (if) resuscitation "restarted your motor", your brainwaves would come back on their own. Why keep a reference for them?

But plastination assumes from the start that revival consists of scanning your brain in and emulating it. Reconstructing you would surely be done computationally, so any source of information could be fed into the reconstruction logic.

Ideally the plastinated brain would preserve all the information that is you, and preserve it undistorted. But what if it preserved enough information but garbled it? The information got thru but it was ambiguous. There would be no way to tell the difference between the one right answer that reconstructs your mind correctly and many other answers that construct someone or something else.

Having a reference point in a different modality could help a lot. I won't presume to guess how it would best be used in the future, but from an info-theory stance, there's a real chance that it might provide crucial information to reconstruct your mind correctly.

And having an external reference point could provide something less crucial but very nice: verification that the process worked.

EEG contains a trivial amount of information, probably not worth storing.

The ratio of the information it adds relative to the total available information is not the point. It's a separate modality. It's subject to a different set of noises and systematic distortions.

I wonder how current cryonics organizations will adapt if/as plastination becomes viable, especially if plastination does well on the critical scanning demonstrations where vitrification is largely unproven.

Can cryonics orgs like Alcor adapt quickly? What about those currently paying insurance premiums based on higher vitrification prices?

Exciting times regardless!

This isn't a new idea, but good to see it getting some press and testing (especially to compare against cryonics). I was initially enthusiastic about the concept as a cheap replacement for cryonics, but currently I favor cryopreservation or possibly a combination approach as the more likely winner.

The main problem (my layman's understanding) is that you can't fix lipids with the cheap relatively harmless kind of fixatives, like gluteraldehyde. Instead you need expensive, hazardous stuff like Osmium Tetroxide. Aschwin de Wolf has written about this in the 2009 issue of Cryonics Magazine.

Overall impression: This is a fascinating and highly valuable research project, but it is too soon to consider it a viable consumer option on par with vitrification yet. My current guess is the cooling + aldehyde partial fixation combo approach may be better for consumers than room temperature aldehyde + OsO4 + whatever else in the near future.

The one thing that makes me think chemopreservation has a lot of potential for rapid improvement is the point Robin makes about mainstream science already having a strong interest in it. Given that, it might improve much more rapidly despite being a technically much harder issue.

I think that it's at least as important, and am backing that up with money (only $100 for now, but I may add to that number later if my financial situation improves). Cryonics requires you to be kept cold in an uninterrupted fashion until you are revived, whereas plastination would allow for cheap and easy storage in a basement. There's the disadvantage that the hardware becomes useless, but it doesn't look like even modern technology is far from making that a moot point.

Exciting news! In my mind the financial failure of a cryonics corp is a very significant factor lowering the probability of cryonics working. If plastination works, it could be a big step forward.

If you are interested in doing cryonics, donating to this group may actually have a positive private return, as plastination is likely cheaper and more effective.

Need to make sure you are preserving the gates in membranes, ion channels, etc (if you don't, likely you won't be preserved). You are probably not preserving microtubules and proteins attached to them and this might matter. DNA methylation and proteins attached to DNA may also be important. I do like the approach as tissue sample preparation though.

It seems to me that plastination is unlikely to preserve chemical details of synaptic protein expression. We really don't know how crucial that is to personal identity, but I suspect it might be possible to have a two stage process, in which the brain is first flooded with dyes, which diffuse into the neural tissue, and bond onto specific synaptic proteins. Then, when the brain is preserved in resin, you have gross visible records of neural and synaptic chemical details.

On a related note, might it be possible to safely break the brain down into more manageable fragments for preservation, by cutting through the sections of the brain that are architecturally simplest, or vary the least between individuals? The corpus callosum and lower mid brain seems like a good place to start.

How big is the "must keep the original atoms" crowd in cryonics? At least Robert Ettinger seemed to think uploading is nonsense because atoms. I'd assume they would not get behind plastination, since it basically assumes viable uploading as a way to get the plastinated brick of a brain into doing anything ever again. Given how many unknowns there are in both approaches though, what people actually get behind might be motivated more by tribal thinking than philosophical first principles.

This is a good point. Ettinger and a few other cryonicists I've talked to are against uploading. My non-serious estimate is that this is perhaps half of cryonicists (who take a side), I'd guess more than half of CI and less than half of Alcor. On the other hand, many others like Mike Darwin seem to be more agnostic on the topic than anti-uploading or the reverse.

It's worth considering that keeping (significant numbers of) the original atoms is not necessarily impossible with fixation techniques, assuming the fundamental ceiling on nanotech isn't too low. Plastination might be disfavored on the grounds that it replaces lipids (old atoms) with plastics (new atoms), but assuming the philosophical attachment is mostly to the proteins you could consider this a viable form of survival as long as the process can be reversed by some kind of sufficiently high grade nanotech.

Fixation in Osmium Tetroxide or something similar could preserve the lipids directly, and might be more along the lines of what the prize competition is doing since plastination is actually something used for art shows, not for electron scanning.

It's worth considering that keeping (significant numbers of) the original atoms is not necessarily impossible with fixation techniques, assuming the fundamental ceiling on nanotech isn't too low.

And that the idea even means anything.

[-][anonymous]12y10

I don't really care if it's the same atoms. Considering that I care about the well being of people that even vaguely remind me of me, and I seem to care more about people I perceive as being very similar to myself, I don't think it's the atoms so much as the "like-me-ness" that I care about.

[This comment is no longer endorsed by its author]Reply

At least Robert Ettinger seemed to think uploading is nonsense because atoms

At least uploading actually does use different atoms. They are still made of the same nucleons, though. A more reasonable complaint would be that it's nonsense because neurons/chips.

Is this an actual argument that people who take cryonics seriously seem to be making regularly though? Sounds like something like Searle's stance, but my unfounded initial assumption is that the crowd that takes Searle seriously and the crowd that takes cryonics seriously don't overlap much.

Is this an actual argument that people who take cryonics seriously seem to be making regularly though?

It's common enough that if you go to one of the Reddit pages for Hanson's post, you'll find someone objecting to plastination over cryonics on the grounds that uploads are about all that one can do with such a brain. Well, yes.

I'll admit, I personally find the anti-upload area in cryonics to be absurd - seriously, you're into cryonics, whose entire rationale is information-theoretic, and you're objecting to uploads? But I have no hard statistics on, say, how many signed up Alcor or CI members are anti-uploading besides Ettinger.

you're into cryonics, whose entire rationale is information-theoretic

I don't think this is the case. Self-identification with your own body can be a strong part of this. I for example personally have a deep emotional connection to my body to the point where I'm much more inclined to do something that has a chance of keeping my brain intact in roughly the same form than an uploaded scan.

Was asking about DanielLC's alternative "neurons/chips" argument, though I'm still not quite sure what its exact content is. Most of the anti-uploading arguments I've seen look like they're either explicitly or implicitly on the "different atoms" grounds. Don't recall many that argue for a fundamental unworkability based on some more sophisticated isomorphism failure.

[-][anonymous]12y40

Has Hayworth or anyone else written up a layman-accessible, or even technical, outline of why plastination might beat cryonics in terms of information-preservation?

Some of the links/citations in my short essay http://www.gwern.net/plastination may be of use.

It's not especially relevant, but the setup of this reminds of the Spanish Prisoner scam.

DEAR SIR:

Good day! I look forward to doing business with you! I am a medical researcher who has unlocked the secret to informational immortality. However, in order to prove my technique works and collect a sum of $100000 (one hundred thousand) USD, I must first raise $25,000 (twenty five thousand) USD. If you can help me by sending me this small sum, I will be happy to provide immortality to you and your suffering species.

Looking forwards to strong business relations, etc.

You'll be horrified when you discover Kickstarter.

Anyone know how this research might be relevant for uploads?

/my current layman understanding

The current most plausible path for uploads seems to me to be brain scanning, which currently operates basically by a sort of plastination process, then slicing the brain into extremely thin slices, imaging each slice at very high resolution, and then assembling it in a computer into something meaningful.

The imaging part is currently in pretty good shape, even if you need to mess around with toxic metals to get something you can use an electron microscope on, and the slicing part has been automated too; one of the remaining bottlenecks is, well, how to process the brain to get it into a slicable state in the first place. This is a question that plastination quality bears upon.

So the only step in the process that needs to be done now instead of post-singularity is the one we're worst at?

Look on the bright side - the only step we need to research now is also the one with the most low-hanging fruit! :)

[-][anonymous]12y10

I think this is great, the hardest part is something we can test now when we know people are interested- and we can probably get things close to perfect with animal experiments.