David Matthewman on the Whole Brain Emulation roadmap

By far the best technical objection I’ve heard so far is David Matthewman’s comments here, discussing Bostrom and Sandberg’s Whole Brain Emulation: a Roadmap:

[…] I still wouldn’t get your hopes up. As far as I can see, it offers no way to discern the state of the neurons, and admits that while it might be possible to get the structure for a small slice of the brain, getting it in 5nm detail for the whole volume is currently impossible, with no known way to overcome the current technological limitations. Many of those limitations are imposed by the wavelength of the medium you’re scanning with, and there’s just no easy way round that. The speed of scanning (which is also currently a showstopper for the ~5nm technologies that might otherwise be attractive) might be able to be improved, but bear in mind that you’re working at levels where the energy of the electrons/photons that you’re using to scan risk damaging the sample, and using more of them in parallel may damage it more. The data transfer/storage problem probably is solvable, by contrast.

I find it a bit worrying that the most promising technologies in the table on page [53] — SOM and SEM, especially combined with array tomography — have relatively little discussion in the text that I can see. This makes me suspect that they’re only even superficially attractive because not enough is known about them to know they don’t work.

Also, given that the conclusion says ‘this sets a resolution requirement on the order of 5 nm at least in two directions,’ there’s far too much discussion of technologies that can only scan down to resolutions two orders of magnitude higher than this. So the text gives the optimistic prediction that ‘[KESM] enables the imaging of an entire macroscopic tissue volume such as a mouse brain in reasonable time’, but what good is that given that KSEM only scans down to 300nm x 500nm? It’s an obvious question, and I’d expect an honestly-written paper to answer it. Because this paper doesn’t, I smell a rat (or, more likely, someone clutching at straws).

The discussion starts ‘As this review shows, WBE on the neuronal/synaptic level requires relatively modest increases in microscopy resolution…’ which may be technically true but vastly understates the difficulty of increasing the resolution of the techniques discussed.

Again, though, I’ll defer to someone who’s done this stuff more recently than I have (and in a medical area — I was mostly looking at metal-matrix composites rather than anything organic).

This stands out from the field in that it is actually in reply to something that someone who believes in cryonics has actually said; it only doesn’t meet the criteria that I asked for in my open letter in that it is blog comment rather than an article, but knowing how busy David is it’s hard to imagine him finding the time to rewrite it in article form any time soon, so with his permission I’m posting it as is.

Updated 2010-02-20: Liam Proven steps up to the plate and meets three of my four criteria. Thanks Liam!

Add post to: Delicious Reddit Slashdot Digg Technorati Google
(already: 3) Comment post

Pingbacks

24.02.2010 11:32 Doug Clow on the Whole Brain Emulation roadmap @blog.ciphergoth.org
[A guest post from Doug Clow. This was a comment in this article on Bostrom and Sandberg’s Whole Brain Emulation: a Roadmap, but given its length and substance I am with permission putting it here as a new blog post.]

Comments

20.02.2010 14:40 Paul Crowley

Having described this as the best objection, on re-reading here I’m not sure I see its force. The paper is trying to be a fairly comprehensive survey of the foreseeable possibilities, so it covers a variety of scanning technologies, but points out where their shortcomings may leave them unsuitable for the task; the discussion of KESM makes it clear that it could not be a complete solution because of the wavelength on which it operates, just as you discuss. And I find over six pages of discussion of EM-based scanning techniques, most of which AFAICT seem to be forms of SEM.

However, I defer to your expertise on electron microscopy — any gaps in their discussion will be much clearer to you than to me.

22.02.2010 6:53 David Matthewman

Just to be clear on the scope of this: no matter how many times you ask me for ‘anti-cryonics’ writing, you’re not going to get it from me, because I’m not ‘anti-cryonics’.

This rely was a specific response to your assertion that: “…it’s plausible to imagine that it might be as simple as serially slicing the brain open and scanning the surface with an SEM and various technologies to find out about the chemistry at the exposed surface, then doing a whole-brain-emulation on the result.” You then linked to this paper as evidence that it might be that simple. I think that paper is strong evidence that it won’t be that simple, and those are my reasons.

Criticising me for (AFAICT) correctly reading the paper, especially as you continue to cite the paper as a valid ‘roadmap’ seems odd. As I’ve said elsewhere, if it is a roadmap at all, it is a roadmap of our current position in the swamp, with big ‘X’s marking all the surrounding alligators. I’m not saying such a map isn’t useful in avoiding getting killed, but until someone invents a completely new scientific technique that allows us to build a bridge, it won’t help anyone get home.

22.02.2010 8:51 Paul Crowley

Just to be clear on the scope of this: no matter how many times you ask me for ‘anti-cryonics’ writing, you’re not going to get it from me, because I’m not ‘anti-cryonics’.

Understood, and I don’t mean to mis-characterize you. Is there a better word for what I’m looking for? Roughly, it’s authors sounding a note of caution and who aren’t already in the cryonics “camp”…

23.02.2010 14:24 Doug

I too am short of time, but have given this paper a quick run through. Here are some unstructured and unedited quick notes I made while I was at it. Apologies for brevity and errors — I almost certainly missed some of their points and have misrepresented parts of their case.

It does seem to be a serious and reasonably well-informed piece of work on speculative science and technology. Emphasis on the speculative, though — which they acknowledge.

The distinction between emulating a brain generically (which I reckon is probably feasible, eventually) and emulating a specific person’s brain (which I reckon is a lot harder), and emulating a specific dead person’s brain (which I reckon is probably not possible), is a crucial one. They do make this point and spell it out in Table 1 on p11, and rightly say it’s very hard.

p8 “An important hypothesis for WBE is that in order to emulate the brain we do not need to understand the whole system, but rather we just need a database containing all necessary low‐level information about the brain and knowledge of the local update rules that change brain states from moment to moment.”

I agree entirely. Without this the ambitious bit of the enterprise fails. (They make the case, correctly, that progress down these lines is useful even if it turns out the big project can’t be done.) I suspect that this hypothesis may be true, but we certainly need to know a lot more about how the whole system works in order to work out what the necessary low-level information and update rules are. And in fact we’ll make interesting scientific progress – as suggested here – by running emulations of bits of the brain we think we might understand and seeing if that produces emergent properties that look like what the brain does. Actually they say this on p15 “WBE appears to be a way of testing many of these assumptions experimentally” – I’d be a bit stronger than that.

Table 2 on levels of emulation makes sense. My gut instinct (note evidence base) is that we will need at least level 8 (states of protein complexes – i.e. what shape conformations the (important) proteins are in) to do WBE, and quite possibly higher ones (though I doubt the quantum level, 11, is needed but Roger Penrose would disagree). Proteins are the actually-existing nanobots that make our cells work. The 3D shape of proteins is critical to their role. Many proteins change shape – and hence what they do or don’t do – in to a smallish fixed number of conformations, and we already know that this can be hugely important to brain function at the gross level. (E.g. transmissible spongiform encaphalopathies – mad cow and all that – are essentially caused by prion proteins in the brain switching from the ordinary shape to the disease-causing one.)

The whole approach is based on scanning an existing brain, in sufficient detail that you can then implement an emulation. I think that’s possibly useful, but I think a more likely successful route to a simulated (!) intelligence will be to grow it, rather than to bring it in to existence fully-formed. By growing, I mean some process akin to the developmental process by which humans come to consciousness: an interaction between an environment and a substrate that can develop in the light of feedback from that environment. But based on their approach, their analysis of technological capabilities needed seems plausible.

The one that leaps out as really, really hard (to the point of impossibility in my mind) is the scanning component. There is the unknown of whether the thing is doable at all (what they call scale separation), which is a biggy, but falsifiable by trying out experiments in this direction.

They talk about electron microscopy as being the only technology which offers sufficient resolution. They say that the neuronal/synaptic level would only require trivial increases in microscopy resolution. That’s missing the point. You just can’t scan enough of a brain with the sorts of microscopy that work at that resolution.

(This is leaving aside the entirely non-trivial question David addressed of whether it’s possible or not to extrapolate from existing technologies to ones that would have sufficient resolution for tiny prepared samples.)

Almost all forms of microscopy that could conceivably come close to being useful here give you an image of an exposed surface. You’re going to need to chop the brain up in to fragments that are thin enough to expose every single synapse, at a minimum. That’s not feasible without destroying at least half of what you’re trying to analyse. When you slice something, you basically smash up a thin column of stuff in the path of the knife (or laser beam, or whatever). And even if you invented some magical way of preparing the samples without that mechanical damage, you’d still have to pull the network of synapses apart in order to expose the surfaces to microscopy .

Sure, you can do it for very small, thin organisms (they mention C. elegans) – because they’re not much more than a couple of neurons thick. But human brains are a lot thicker than that. And anyway, C. elegans was (I strongly suspect) done by scanning loads of individuals and aggregating the date. That’s automatically shut you out of the big-ticket replicating-a-person bit.

Oh, and it’s worth noting that all of this, to be remotely possible, is really quite spectacularly destructive. Your brain is not going to be doing any thinking once this process is done with. Which is another reason why I think growing a simulated/emulated brain is a better research plan.

There are some imaging techniques which don’t require the slice-and-dice bit: MRI is a better possibility, at least superficially. And this is probably something I could dig in to more later, since I do know quite a lot about the physics/chemistry behind the technique. (Part of my PhD was developing a teaching simulation of an NMR spectrometer, which is a simpler thing than an MRI machine.) But off the top of my head it really doesn’t seem likely for all sorts of reasons — resolution, in multiple senses (you’re going to need a radio sensor with finer resolution than is theoretically possible). If you want, nudge me and I’ll try to find time to spell this out more and may even dredge up enough of the maths to do sums on it.

Their ideas on nanodisassembly seem like nonsense. You can’t build Drexler-type nanobots: the physics/chemistry just doesn’t work like that at that scale. Think proteins, not a teeny version of Robosapien. They say (p51) “Given that no detailed proposal for a nanodisassembler has been
made it is hard to evaluate the chances of nanodisassembly”. I don’t think it’s hard to evaluate: the chances are negligible to nil.

Chemical analysis – this is really not going to happen. I’d just been thinking about neuronal connections. Blimey, they are really stretching the bounds of what’s even theoretically possible here. There’s several techniques they mention that I don’t know about, but I do know about some and SFAICT they all suffer from the need-to-break-the-brain-up problem only worse. They mention dyes – but dyes are generally pretty large in chemical terms and will almost certainly destroy information if you perfuse a brain with them.

I’m not paying any attention to the information processing stuff. I could do, and the challenge is large, but (a) my seat-of-the-pants feeling is that Moore’s Law can do more than enough here, and (b) lots of people – ciphergoth included! – are at least as capable as me of doing the detailed scrutiny here.

Likewise the image processing and scan interpretation bit, and the neural simulation component, and so on – that doesn’t strike me as the really theoretically hard part. It might turn out to be practically impossible, of course, but it doesn’t ping my bogosity meter the way the scanning part does.

Ah, actually, one thing on the implementation side that might scale up to be infeasible is if you need to get serious about shape variability of proteins. Calculating possible conformations of proteins is (currently) a classic application that requires Grid computing type resources: it’s at the edge of what we can do with current computing resources. And that’s simply twisting a single smallish protein around. If we have to do that for a large proportion of the proteins in a brain, it’s easily above what’s going to be computationally feasible this side of the singularity. But I suspect we’ll be able to get by with a lot less detail, though not nothing — e.g. for TSE prion proteins, we might not have to do the whole calculation of possible conformations for each protein, and how that’s affected by its neighbours: we might well be able to just model it as a location in space and a one-bit state variable indicating whether it’s in normal or TSE conformation. My guess is that you’d need a bit more than that but not a lot. But that really is a guess that wants empirical testing.

23.02.2010 14:43 Paul Crowley

Wow, thank you very much. Do you mind if I turn this in to a post in its own right? I believe I can simply mark your account as having posting privileges so that it can directly be in your name. Or of course if you prefer to post this to your own blog that would also be wonderful!

23.02.2010 16:24 Doug

No, feel free to bump it up if helpful — and happy for you to do it i.e. I haven’t time to :-)

Not really on topic for my work blog and my LiveJournal isn’t (supposed to be) search engine-able so better for your purposes to put it here and I’m happy to have it done that way.

For the avoidance of any doubt, have a CC0/public domain release: To the extent possible under law, I hereby waive all copyright and related or neighboring rights to the blog comment I made above, which I made from the United Kingdom.

Comment post