Aeon Sub-Title: Your body is scanned, destroyed, then reproduced. Do ‘you’ live on the copy?
- This paper reviews a Closer to Truth interview of Andy Clark by Robert Lawrence Kuhn that appeared on Aeon1 on 19 August, 2019. There are probably clearer and more extensive accounts by Clark on this matter, but it’s interesting to see him in action. He seems very smart and confident, if utterly misguided!
- Note that while the sub-title might appear to be tendentious – in saying that the reproduction of the destroyed body is a “copy”, as it undoubtedly is – this is only a problem for your survival if you are numerically identical to your body, which many deny.
- Closer to Truth has started to provide transcripts, but not for this interview, so – after noting the Aeon summary – I provide a summary of my own, followed by a commentary. I’ve included the YouTube timings to make it easier to match my paragraphs up against the video itself.
- I’ve put a fair amount of work into this, so the lessons should be summarised. My overall opinion is that Clark’s initial intuition2 – the “we are patterns in information space” – raises lots of interesting questions, and gets round most of the difficulties associated with taking personal identity to be subject to the laws of classical logic3, being an equivalence relation. But I just don’t share his intuitions. We just are individual things, and these things have experiences that are important to them. The experiences themselves – or the information that is contained in them – aren’t us4, but are properties5 of us.
- That said, there’s a couple of items I’ve recently come across, namely:-
→ "Han (Byung-Chul) - The copy is the original", and
→ "Han (Byung-Chul) - Shanzhai: Deconstruction in Chinese"
that beg to differ – at least as far as artifacts6 are concerned. Additionally, the author points out that artifacts can be maintained, much as organisms7 maintain themselves.
- For centuries, philosophers – and more recently, science-fiction writers – have been concocting riffs and variations on a particular thought experiment: if every bit of your body could be perfectly scanned and replicated, in what ways would the replica still be ‘you’?
- In this interview from the PBS series Closer to Truth, Andy Clark, a professor of philosophy at the University of Edinburgh, dissects a version8 of this experiment posed by9 the US philosopher Daniel Dennett, in which a body is scanned, destroyed, and replicated in a distant place.
- While science hasn’t yet brought us close to putting Dennett’s conundrum to the test, we can still grapple with the intriguing and perhaps troubling metaphysical questions it raises, questions that might become even more material as we careen further into the information age, including:
- Would ‘you’ be dead, or would your sense of self perpetuate in the copy?
- And, if you were recreated several times, where exactly might you expect to find your embodied sense of self?
- Kuhn (00:00): There’s an assumption – a “given” – among technologists and futurists that it’s only a matter of time before the “engineering problem” of how to upload our subjective consciousness to a non-biological device is solved. It’s an “engineering problem”, and while 10 years is wildly optimistic, no-one in this camp doubts that success in hundreds of years is conceivable. Kuhn doesn’t think this is intuitively obvious, and asks whether Clark – with his background11, knowing what consciousness is and what it does – does think so; so that his first person consciousness can be uploaded so that – in a sense – he’ll feel no different than he does now.
- Clark (00:55): There’s something in Kuhn’s set-up that makes him uncomfortable. The notion that there’s a consciousness here that might be uploaded over there. Where he wants to start is by thinking of “my consciousness” – myself – as a pattern of some kind: a pattern in information-space. If you stop thinking of yourself as a locus of this special thing “my consciousness”, but think of your consciousness as a pattern of some sort, then the concept of uploading doesn’t seem too outlandish. The upload itself would be “kinda weird” because it would just be reproducing that pattern rather than pushing it down a tube.
- Clark (01:45): There’s a nice little thought experiment that originated with Daniel Dennett called “Death or Transport”. Brief description: Technology exists where you get into a cubicle in one location, your body is scanned and the information about how everything is organised is transmitted to a distant location, and the result of the scanning process is that the original body is destroyed. At the distant location, the pattern of stuff is put back together in the same kind of way and something that looks and sounds like you gets out the other end. Dennett’s question is “is this death or is it transport?”. Would you get in there and would you think this is a way of getting from A to B safely or not?
- Kuhn (02:45): Well, if we say “that worked”, since we have “the information” we could repeat the process – maybe multiple times – and we have a multiplicity of Andy Clarks all saying “that worked”. But then each one would be shocked to see one of the others.
- Clark (03:15): That’s true – though I might be less shocked than some people. Some people would be outraged – John Searle, for instance – because these (teletransported) people are just pretending to even have consciousness, let alone being you, depending what you made them of.
- Kuhn (03:35): But there is a thing to be your internal consciousness right now – I can feel myself being that – and when we go into this multiplicity, even though every one of them will claim it, will any of them be it?
- Clark (03:55): I think that’s a very funny question. For example, suppose that we just froze you completely somehow – so there’s no neuronal activity whatsoever – and had the ability to bring you back again. I think most of us would be fairly happy with that. Most of us would think that a little interruption in the pattern that is me doesn’t stop it …
- Kuhn (04:20): General anaesthesia … you get it … you’re out …
- Clark (04:25): But then there’s certain subterranean processing and stuff …
- Kuhn (04:30): But then … I come round and I know it’s the same me ... but I feel that there’s a qualitative difference between that and uploading it in an informational form in a different locus.
- Clark (04:50): But then if we are all patterns in information space, then the locus isn’t an essential property of “you-ness”. I don’t actually know how to resolve this question, so at this point someone with other intuitions – someone with the “me, here, now, animal” intuition – is just going to be led to think about the world in very different ways than I do. One of the implications of thinking about persons as patterns is that all these questions about personal identity don’t seem so “all or nothing”. There’s a nice Sci-Fi series on British TV called Black Mirror12, and one of the stories there was a near-future in which our social-media records have become so detailed and intelligent that social media could start putting out stuff as if you were putting it out after your biological demise. Of course, if you think you’re a pattern in information space, that might actually be a way of persisting as a shallower version of yourself, and if there was enough of a pattern there, I’d have to say that just is you. So it’s because of this almost metaphysical presupposition that I’ve got that we’re patterns in information space that notions like uploading just don’t seem that difficult. But I don’t know how to resolve this question – how do we know that we’re a pattern in information space or a pattern in meat space?
- Kuhn (06:25): If I give you that we’re a pattern in information space and if you have ten versions of yourself that are – at this microsecond – exactly equal: from then on they would differentiate, but at that moment each one claiming to be you … you wouldn’t say that there’s a fractionalised consciousness … that they’re all part of the same consciousness? Or you wouldn’t say one is real and one is not real. So, you’d have these ten separate centres … but right now you don’t feel ten separate centres – you feel like one – so would you be one of those ten, or is it an irrelevant question?
- Clark (07:10): I have to say I would be all of them. But it’s not as though there would be one consciousness. I think consciousness exists at the meeting point of perception and action. So, for each of those creatures there’s a sort of the meeting point of perception and action and their consciousness is constructed in that meeting. That’s why I don’t think it’s like a cloud that suddenly has all ten of us in – but there are ten “me’s”.
- Kuhn (07:45): Each of the ten are going to assume that they are the original that got into the transporter and will be shocked to see that there are others.
- Clark (07:57): Yes; certainly if there are ten locations, that would be true and even if there is one location, the one that gets out at the other end that expect the meaty thing at the other end to be destroyed will be shocked to find it is nonetheless still in existence.
- Kuhn (08:15): They would think that the one that they are seeing back there is some sort of an imposter, because they knew – remember – that they got in and are now here. They won’t give credibility to that back there.
- Clark (08:35): I’m not sure about that. Remember they got in knowingly – that is, you’re already the kind of person who thinks that you’re a pattern in information space, else you wouldn’t have got in in the first place. So, when they get out the other end … if it turns out the machine has malfunctioned and the meaty thing is still in the cubicle, then they are going to think “Oh dear, I’m in two places” … I didn’t expect this, but we are going to have to come to some sort of amical arrangement about things here. It’s just in fact … suppose there were ten of these – it doesn’t matter how many you have – that they would come to think that some sort of priority should be given to the original meat version that got into the scanner – maybe that one gets the car. I think that if you think of yourself as a pattern in information space you’ll think these are the kind of issues that we’ll ultimately have to learn how to resolve.
- Kuhn (09:38): And you would be willing – assuming it were proven by various forms – to give it a shot?
- Clark (09:45): I would want to share everything with the others actually because I believe that I am all of the others. So, in that sense I think I would …
- Kuhn (09:50): But we’re not talking about a fractionalised consciousness or a split personality. Each one is a total independent consciousness just like we are today. But still, I’m torn by this fact that you have in your mind this inner sense of who you are and in radical distinction from everything else – from every person or object – you have that inner sense. Where would that go – would it go to all of them?
- Clark (10:30): It wouldn’t be shared amongst them all, but they would each have that sense that they are a locus of consciousness and they would be right.
- Kuhn (10:40): Sure they would, but the original that walked into that machine – would that have disappeared? The meat that walked into the machine itself?
- Clark (10:50): Well, currently we’re assuming it didn’t disappear, right?
- Kuhn (10:53): Whether it did or didn’t disappear it’s the same question. It’s easier if it didn’t disappear.
- Clark (11:05): So, now you’re beginning to think it’s death … heading toward that because there’s this sort of persisting locus through time and once that’s gone, even if the pattern re-emerges it’s not quite you that’s re-emerged. But I think that the right sort of consideration is that of how you’d feel if your pattern was temporarily interrupted as by being put into deep refrigeration or perhaps even during deep sleep.
- Kuhn (11:39): The way I’d look at it is that my relationship to those others is as it would be to an identical twin brother where we have the same DNA but are totally different people.
- Clark (11:52): Well, this would be a really, really, really identical twin. We don’t have experience of things like this, because there are differences between those twins and from the get-go, of course, their life histories diverge.
- Kuhn (12:08): But the life histories of all the ten will begin to diverge …
- Clark (12:10): They will, and in ten years’ time they’ll be …
- Kuhn (12:15): What do you think is the likelihood of this actually occurring will be, given enough time?
- Clark (12:20): I don’t see anything like “death or transport” technology in the next hundred year, and wouldn’t like to look further. What I do think we’ll see is more and more technology preserving elements of the patterns I would think of as us. So, for example in social media and in other possible forms. We’ll have to confront this question – to what extent are the thickest versions of those preservations actually sort of versions of us or not. Is this a little tip-toe into the sort of realm of virtual immortality or virtual regeneration? Or is it just like a few lines left over in a notebook after we’ve died?
- So, the TE14 under consideration is of uploading “your consciousness” – throughout presupposed to be “you”, thereby also presupposing some version of the PV15 – to a non-biological device, presumably a digital computer of some sort. The question is “how this feels”, “feeling” not being something that computers are noted for, rather than some more cognitive question. Note that this is – on the face of it – not the same TE as Teletransportation16, though Clark raises this TE in the third section. Uploading17 is the hope of transhumanists18 as a means of achieving “virtual immortality”. This phrase – the title of the Video – is ambiguous, maybe intentionally. Firstly, the immortality is not truly thus, as it depends on the survival of the computer, or its descendants, which not likely given the likely ultimate heat death of the universe. Secondly, it is immortality in the virtual world. The big question is whether this “immortality” is a mere simulation.
- Clark is right to say that consciousness19 isn’t a fluid that can be squirted up a tube. I would say that it’s a property of some other object – I have consciousness, or – putatively – the computer (or a portion of it) has consciousness. Clark, however, thinks that his consciousness is (numerically identical to) he himself, rather than a property of himself. His primary intuition20 is that it is a “pattern in information-space”, but he doesn’t explain precisely what this is in this video. But, he’s right to say that if this is what it – and he – is, then it makes sense to say it can be uploaded – ie. copied – into a computer. I would just note here that – as will become apparent – numerical identity goes out of the window on this account, though it might be saved on a 4D (eg. Perdurantist21) account of persistence.
- This is a straightforward description of Teletransportation22. Click here for references. In itself, this is only relevant to the case to hand insofar as we’re reducing persons to information, at least for purposes of transmission. This information is not uploaded to a computer, but used to build another human animal. Clark attributes this TE to Dennett, and he is correct, as "Huenemann (Charlie) - If I teleport from Mars, does the original me get destroyed?" points out. It’s in the introductory paragraphs of "Dennett (Daniel) - The Mind's I - Fantasies and Reflections on Self and Soul: Introduction". However, I think its first fully-worked-out case is due to Parfit. Wikipedia cites "Parfit (Derek) - Divided Minds and the Nature of Persons", and the section “What We Believe Ourselves to Be” does cover it, though I think "Parfit (Derek) - What We Believe Ourselves To Be", from "Parfit (Derek) - Reasons and Persons", is earlier. Anyway, would this – as Dennett asks – be death or transport? My view is that it’s – as Dennett’s naming of the machine as the “Teleclone” suggests – a case of cloning23.
- Kuhn doesn’t directly answer this question – other than by implication – but raises the reduplication objection24: information can be “realised” multiple times. His comment that the duplicates25 would be “shocked” at the existence of the others is irrelevant. As Clark notes in the next section, some will be more shocked than others. The real issue is what this says about the logic of identity26.
- Clark says that some would be outraged, but not in the way that others would be “shocked”. The “outrage”, though he doesn’t explicitly say this, would be caused by the misleading description of the result of the device as “transport”. He reverts to the original TE of a computer upload by saying that John Searle would deny that the Teletransportees even have consciousness, let alone the “same consciousness”. Searle wouldn’t deny consciousness to traditional “meat-based” Teletransportees, only to digital-computer uploads.
- Kuhn points out the logical distinction between seeming to be something and actually being it – really the distinction between the original and an exact copy, the distinction between numerical and qualitative identity. He evidently thinks – as I do – this matters.
- Clark doesn’t understand things that way at all. He describes Kuhn’s question as “very funny”, while making an analogy I find “very funny”. Clark seems to think the situation comparable to a human being being frozen and then thawed out (ignoring technicalities) – while there is a temporary interruption in the “pattern” of consciousness, most would be happy that identity is preserved. They would, but the situation isn’t comparable, in my view.
- Kuhn suggests something less sci-fi – general anaesthesia – on the presumption that this also provides a complete cessation in the pattern of consciousness.
- Clark suggests that there would still be unconscious mental processes going on. If this is relevant, then Clark’s view of what we are includes unconscious mental processes – as it ought to do.
- Kuhn thinks there’s a qualitative difference between just “coming round” and being uploaded “in a different locus”. In the former case he knows “it’s the same me”. How? He doesn’t say, but it can’t be from first-person evidence – what it seems like – because the teletransportees also, from a first-person perspective27 think they are me. I’d be interested to know whether Lynne Rudder Baker agrees. I think her view – in the case of fission – was that there was a “fact of the matter”, but I’m not clear what that might have been. Anyway, the evidence that it’s him that wakes up must be based on other factors (same bed, common practice, 3rd-party evidence that he’s not been swapped out, etc.).
- Clark counters that if we are “patterns in information28 space”, then the locus isn’t an essential property of you-ness. But he doesn’t know how to justify this intuition29 as against an animalist intuition. See my Note (just referenced) for whatever thoughts I might have on this question. He makes an important point about the “all or nothingness” of PID statements. He rejects this dichotomy, as did Derek Parfit, and – maybe – for the same reasons, though Clark doesn’t mention Parfit. But, he must think that Identity is (not) what matters30 in our persistence, but simply matters of psychological continuity and connectedness31 – and so persistence is a matter of degree. I thought the Black Mirror TE deeply counter-intuitive. I can’t see how any “pattern in social media”, however detailed and however it might impersonate “the deceased you” could actually be you.
- Rather a straggle of possibilities raised by Kuhn. In the classical logic of identity, there do seem to be only three options in the case of “ten for one”.
- Collectively, the ten are you: on the PV32, there’s “distributed consciousness”, though as the ten individuals are isolated, it’s difficult to see how this would work. On a non-PV, the ten might be a scattered object33.
- Only one is you: this is a rather unprincipled version of the closest continuer34 view.
- All are you: this would require a version of 4D (either perdurantism35 or exdurantism36).
- Clark confirms that there would be ten of him, and there wouldn’t be a scattered consciousness, since “consciousness exists at the meeting point of perception and action”, so each of these individuals would be individually perceiving and acting. He doesn’t mention 4D, and that’s – I presume – because he takes himself to an abstract object rather than a concrete particular. What else could a “pattern in information space” be? Maybe he thinks of himself as an instantiation of an abstract object – like a copy of a book, though a copy that can evolve over time.
- Kuhn repeats his assertion that each will be “shocked” to find there are others, given that each will assume it is the original.
- Clark agrees. It’s not clear what TE is in mind here. “Ten locations”, presumably refers to the information being transmitted and implemented in ten individuals. But these are contrasted with the “meaty thing” that ought to have been destroyed, but remains in existence, implying that the ten are not “meaty”. So, the ten individuals in different locations may be uploads to ten computers; and “one location” may be ten uploads to different partitions of the same computer? But he does use the expression “gets out the other end”, though this may be in a virtual world, or as some robot not “made of meat”.
- Kuhn claims that because they remember getting in, and have now got out at the other end, they won’t “give credibility” to the one who got in and – contrary to the agreement – hasn’t been destroyed. This seems unlikely to me. It would cause an existential crisis, and bring back long-suppressed thoughts that such methods of “travel” in fact lead to death and duplication37. The “original” would have precedence in any closest continuer38 schema.
- Clark thinks surprise is unlikely, saying that anyone using this form of “travel” must already believe that they are a “pattern in information-space”. And would therefore think that they were in two places, but that the “original” deserves priority. He thinks this sort of conundrum would have to be addressed. Quite so, of course. But I imagine that if this sort of case was very rare (and maybe not publicised when it occurred) and if it had become common wisdom that this was just a form of quick transport – after all it’s been going on for years and people go and come back looking none the worse – then the vast majority of people won’t think of themselves as “patterns in information-space” or anything of the sort. Also, if they thought they were at any sort of risk of having to share everything with duplicates39 (or even other selves) they’d not take part in this game.
- Kuhn asks Clark whether he’d be willing to enter the transporter, assuming it’d been tried and tested.
- Clark doesn’t directly answer the question, though says “in a sense he would” (I’m not sure in what sense) but addresses the consequences of a bungle – in that he’d share his goods with all the others, as he’d be all of them.
- Kuhn presses Clark about (in other terms) the destination of the FPP40. It wouldn’t be shared, so would it go to all of them? This is an issue for Baker41 in the case of fission42. As far as I remember, she claimed that there would be a “fact of the matter”, but I don’t know what determines it, either from a first- or third-person perspective. How would anyone know?
- Clark repeats that each would be right to consider itself a locus of consciousness.
- Kuhn agrees, but asks a confused question about the “meat” that entered the transporter – has it disappeared?
- Clark clarifies the TE under discussion – where the teletransportee still exists.
- Kuhn’s confusion continues – it’s “the same question” whether or not the original disappears. It’s “easier” if it didn’t. I suppose what he means is that if the original survives, the question of wo is who is easier – it’s the original, and the others are copies, but he doesn’t make this point plain.
- Clark now clarifies Kuhn’s thought – that it’s “death, not transport”. It’s the physical locus that’s important for Kuhn, and the re-emerging pattern is “not quite” him. This shows that numerical identity is not what’s in mind from Clark’s perspective, as NI is all or nothing. But Clark asks what Kuhn’s view would be if his “pattern” was interrupted by deep-freeze, or maybe deep sleep. This is an interesting question, but only for someone who holds to the PV43, as then we’re talking about an intermittent object44, and it might as well intermit across space as in the same place, as a mind (or pattern of information) is not spatially located in any case. But, for an animalist45, or even a supporter of an old-fashioned Body Criterion46 or Brain Criterion47, the physical continuity48 is essential.
- Kuhn ignores this suggestion, and says he’d relate to the others as identical twins49. While they share the same DNA, they are different people. This situation isn’t quite parallel, as Clark proceeds to point out.
- The lack of parallelism with the “identical twin” case is that the twins started to diverge physically soon after the point of twinning, and mentally at the point of birth, at the latest. While the “instructions” in the DNA may be identical, their implementation – the epigenesis – is dependent on countless environmental factors. So all identical twins we have experience of aren’t really that identical – certainly not from the psychological perspective (as “patterns in information space”). Clark doesn’t actually point this out in detail, but no doubt has it in mind.
- Kuhn points out that the life histories of the ten will begin to diverge. The conversation is drawing to a close at this point, so this point isn’t really followed up. If all these are “Clark”, how come they are increasingly different patterns in information space. This raises questions about the normal case of the non-fissioning Clark. What makes one “pattern in information space” identical to another, at a later time. Presumably, identity doesn’t come into it, let alone “matter”.
- Clark is cut off mid-sentence, so we don’t know what he thinks the situation would be in 10 years’ time for the ten duplicates50 in this TE, which is a shame as – prima facie – it’s incoherent to say these very different individuals are all him.
- Kuhn: what’s the likelihood of “this” happening … ever? He doesn’t say what “this” is. The original question was on uploading, but shifted to teletransportation51, which is technologically much more difficult, one would have thought.
- Clark thinks it’s at least 100 years for anything like teletransportation52. He seems to provide some support for the – to me, absurd – suggestion that we would live on “in the cloud” if our social media presence were detailed enough. This might be the start of the “virtual immortality” of the video’s title. This seems a reminder of Woody Allen’s quip about not wanting to live on in his work, but by not dying (See my Note on Immortality53).
Footnote 1: Footnote 8: Footnote 9:
- We’re not told where, either in this summary or in the interview itself.
- Also, I don’t think this is originally down to Dennett.
- This is very detailed, because I don’t want to miss out on important points – particularly those I’m not inclined to accept.
- However, it is not always verbatim, though it often is, more or less; but I’ve not tried to capture any rapid interplay between the interlocutors.
Footnote 12: Footnote 13:
- See the links under Andy Clark.
- Kuhn mentions that Clark has studied consciousness, embodied cognition, extended mind, predictive brain, …
- Indexed to the numbering system in “My Transcript”.
Text Colour Conventions (see disclaimer)
- Blue: Text by me; © Theo Todman, 2020
- Mauve: Text by correspondent(s) or other author(s); © the author(s)