In Defense of Cryonics
The current issue of Technology Review contains Michael Hendricks’ scathing condemnation of the proposal of cryonics, i.e. the idea of freezing your body or at least your brain as soon and as rapidly as possible after death, with the goal of being revived in a distant future where technology has progressed to the point where it is possible to do so.
Due to the difficulties of recreating decayed and destroyed tissue, this resurrection might well not be in the flesh, but could involve scanning and recreating the configuration of the brain as closely as possible, using a computational simulation. It is this particular vision that Hendricks argues against. Hendricks is an expert on nematodes, specifically the roundworm Caenorhabditis elegans, which tends to live in soil and feed on decaying organic matter when not spending time in the lab. I can see why Caenorhabditis dislikes the prospect of his food ending up in Alcor’s freezer instead of the soil, but the core of Hendricks’ arguments is not as compelling to me.
Hendricks makes five especially salient points:
1. It is and will always be impossible to recreate the functionality of a brain from its connectome (i.e the map of its neural connections). Hendricks’ has looked at roundworm connectomes, and found that a mere scan of synaptic connectivity is insufficient.
Of course, synapses might be either excitatory or inhibitory, or several other things, in different proportions. Thus, a scan that only tests for the existence and position of a synapse won’t even tell us if a particular connection will activate or suppress a neuron. But what about recording the type and positions of synaptic vesicles (little bubbles with specific signal molecules that determine the interactions of the synapse with the associated dendrite)? Perhaps there are even shortcuts, i.e. markers that allow to predict which vesicles are going to manifest in a given synapse? I do not see a convincing a priori reason why that should be impossible in the future. By the way, neurons are unlikely to tell the whole story; researchers will probably also need to look at glia cells, blood vessels and a bunch of other things we don’t even have on the radar yet. But it is unreasonable to assume that progress in connectomics is impossible, and functionally accurate scans of neural connections will forever elude us.
2. A lot of the non-neural properties of a working brain are currently not anatomically visible, especially the interactions of large populations of neurons with neurotransmitters that swamp a whole brain region.
But if these are global interactions, not targeted at the level of individual neurons, their replication might turn out to be easier, not more difficult than individual synapses?
3. A dead brain lacks the activation and much of the precise molecular configuration that determines a brain state, including memories etc.
But brains should also be resilient against losing that state, so sleep, fever, shock or psychoactive drugs do not permanently alter our personalities. We are sure to lose all memories that have not yet been encoded at the synaptic level, and many more due to decay and freezing damage, but in theory, it is imaginable that a sufficiently advanced brain scan technology could restore a lot of what we were.
4. Even if we could revive a new instance of our brains in the future, this copy would not be “me”, but merely a twin with identical memories.
This philosophical objection is tricky, and I believe it is more wrong than the others, because it assumes that personal identity is somehow an objective feature of the world, which it is not. Personal identity is a construction within a particular mind. Tomorrow I will wake up as a near identical twin of today’s self, with a sense of personal identity being supplied by memories that connect my self of tomorrow with the remembered self of the past. The same would happen to future me, after being scanned and revived. There will be a break in the physical continuity, but none in its internally constructed identity. Future me after revival can be just as happy with being me as tomorrow-after-I-wake-up me. Of course, today’s I won’t be there to join in the fun, but it will also not objectively be there tomorrow morning. If that worries you intuitively, I can offer a psychological explanation to explain your intuition, but I don’t think that you need a metaphysical one.
5. The proponents of cryonics exploit the grief and cluelessness of people that have lost a loved one for economic gain, and therefore deserve anger and contempt.
As far as I can tell, this is not a good description of the state of affairs. The cryonics company Alcor (which does not make profits, but relies on occasional donations by wealthy affiliates to keep it afloat) puts the probability of reviving one of its customers at somewhere between 0.023% and 15%. Unlike the billions of subscribers of Abrahamitic religions, Alcor’s 2000 subscribers are not given a promise of certain resurrection and eternal life. Alcor only offers a small (but non-zero) chance of prolonged existence. Even if Alcor’s probability estimates were off by a couple of magnitudes (have a look if you buy their arguments), it does not seem irrational to me to put some spare $80.000 or so into a slightly eccentric method of burial, if one values their existence highly enough.
I agree with Michael Hendricks that it is dramatically more likely that cryonics is not going to work, and that its current proponents do not have any technology to even partially reconstruct a human nervous system. But unlike Hendricks, I think that the cryonics companies and customers are aware of this, and deserve neither anger nor contempt.
(Disclosure: I am neither affiliated with a cryonics outfit, nor do I currently intend to become a customer.)