Cryptic ontology

Steve Grand said this:

"The mistake, IMHO, is to assume that physical things have some kind of special ontological status - that an electron is a little lump of matter sitting in space and somehow more "real" than, say, a cloud or a whirlpool or the operation of a program. All "things", all "solid objects" are really emergent processes - an electron is a resonant state in the electromagnetic field. So if a simulated electromagnetic field spontaneously curls up into a similar resonant state (as it should if it is a good model) then I think this resulting virtual electron has just as much right to be called "real" as a so-called physical electron does. The properties of the virtual electron are not "programmed in" - they emerge from interactions that are the same as those in the physical world, even though the properties of the medium from which they arise are sham. They just exist in a different universe, that's all."

I essentially agree with this. Like Steve, I don't think it makes sense to distinguish between "abstract objects" and "physical objects" in any way. I think this is different from how most people see the world. Most people would think of a piece of rock, or a dog, as a physical object and would think of a novel, a computer program, or a picture of the Mandelbrot set as abstractions. Both Steve and I tend to think that everything has the same ontological status. We may think we are "physical" but we are just patterns like everything else. Our bodies are no more or less real than a wallpaper pattern. That means that either our bodies are not "physically real" or the wallpaper pattern is physically real. I tend to take the view that the word "physical" does not even mean anything. It would only be of any use if we wanted to distinguish between physical and non-physical, abstract things - and both Steve and I deny they exist.

Does this mean that "we aren't really real?" - I would say, "No!" - it is just a deeper idea of what "real" means. You are just as real now as you seemed to be yesterday.

Steve and I have discussed these issues at some length in e-mail and found we agree on a lot.

There is one way in which I may take a more extreme view than Steve. I do not distinguish between sensible patterns and contrived patterns. As far as I am concerned, any abstraction you could make of anything corresponds to a real object. I am not saying Steve disagrees with this. I think he may be a bit suspicious of it but he has not expressed a final opinion. The consequences of this are extreme. As an example:

The Mona Lisa painting, when subjected to a particular interpretation, in which the colour of each point in the image can be associated with the colours of the points below it to make a grid, gives the Mona Lisa picture. We should therefore view this Mona Lisa picture as a real object. A relatively simple analysis of the painting exposes the pattern.

But:

Consider my desk. By applying a complex interpretation to my desk you could also get the Mona Lisa painting. You would need a complex set of rules which say how to look at my desk and get the Mona Lisa pattern out.

There is no way in which we can say one abstract pattern is more justified than the other - in my view, both are real.

But:

By analysing my desk you could extract any pattern you wanted. You could say my desk is running a chess program if you are allowed to look in a complex enough way. You could, in principle, extract a description of our entire universe, or someone else's entire universe from my desk.

Rather than get stuck in some awkward situation of saying what the reasonable interpretations are and what they aren't I make it simple: all abstract interpretations of any object are real objects.

This is clearly something with a lot of implications. It would mean we are living in a very extreme multiverse, where every object implies the existence of every other object as an abstraction of it. It would also mean that what we think of as "physical reality" is not even physical reality. It is just an abstraction of something else, which is in turn an abstraction of something else and so on.

I wonder what you think of that? I know it will sound extreme but if you want to try to get a handle on it, imagine this game.

Fred has had himself uploaded into a computer. He has himself encrypted to hide. He has himself more strongly encrypted the next day, and the next day still more strongly encrypted. As this goes on it becomes harder to find Fred. The only way you could find Fred is by doing a really complex interpretation of the computer system. Eventually, this interpretation would be so extreme it is even more complex than Fred himself! Has fred ceased to exist? I would say no - as far as Fred is concerned, nothing changes. However, to use Steve's kind of language he has more or less moved "into another universe". He is separated from us not by light years or wormholes but by something more extreme - a kind of gulf of complexity of interpretation. We could say "Fred is hiding in the computer - if you analyse it deeply enough you will find him!" but this may seem an act of faith to some people - and the same reasoning could be used to say that the Mona Lisa, or a simulation of another universe is hiding in the matter of my desk. As far as Fred is concerned, reality is going on as normal and Fred could apply complex interpretations to the structure of his own world that describe other objects, and so on. If anyone objects to this my question is simple: at what point does Fred's encryption become so extreme that he ceases to exist - and why?

Another example:

The government declares all pornography a work of Satan, and people can be imprisoned for possessing it - even encrypted pornography - no matter how strongly encrypted - and regardless of whether they possess the means to decrypt it. Anyone who really thinks about this issue would see that it would make it illegal to own a copy of Microsoft Word, the
Bible, a desk, or a meat pie - because by applying a complex enough logical manipulation to these items you could "decrypt" anything you wanted. Any sensible administration would have to impose some kind of practical limit on what they meant by saying one thing was strored encrypyed in another system - and it should be obvious that such a limit would be arbitrary and not about anything profound.

To anyone who buys into what Steve and I think about patterns corresponding to real objects, and what we think of as "real objects" just being patterns with which we are very familiar, this should be very disturbing. In fact, I would say anyone who does not have a strong argument about why this problem of "arbitrariness of interpretation" does not arise, and who does suddenly sense an ontological abyss before them has not understand what many people think of as quite philosophically awful implications of this.

My view is that the sort of view that Steve and I have, in which there is no ontological distinction between patterns and physical things, inexorably leads to the consideration that you can say that any pattern is in anything if you look deeply enough, and that leads inexorably to a crazy kind of multiverse - not just one in which every possible universe
exists, but one in which every possible universe is implied by every single tiny piece of any other universe, and one in which nothing is fundamental. I am not sure what Steve's view is of these extreme implications, or how he would answer them - but I see no escape from it.

I also think Searle's argument has an important role in this, even though I think Searle is wrong, because his case needs answering. I think Steve answers it properly, in his comment - but, in my opinion, with what some people may regard as very disturbing implications about our place in reality.

I do feel resistive towards your more extreme views, yes. That doesn't mean
you're wrong, just that they take some getting used to. I guess we can
define "real" any way we like, but to define everything in the universe as
simultaneously being everything else seems like overkill to me. Maybe I
still don't understand.

I like your pornography illustration - that gives me something to get hold
of. I agree that you could "decrypt" the Bible and end up with a
pornographic picture. However, as far as I can see, you could only do that
and intentionally end up with a SPECIFIC picture. There will always be some
mapping that gets you from one sufficiently complex object A to another
object B. But I can't imagine any mechanism of any kind that could "decrypt"
a given object A and end up with an unknown member of a CLASS B. There isn't
any way that the Bible could be "revealed" to be pornographic in a general
sense without knowing in advance which specific picture you intended to
prove it was encoding. If your job as a member of the Thought Police was to
prove that a variety of such "subversive" materials were really encrypted
porn you would need a different decryption algorithm (or key) in every
single case. And you would have to try a huge number of keys to find one
that worked. Moreover, those keys would be roughly as complex as the thing
they claim to decode. Isn't that a "practical limit on what they meant by
saying one thing was stored encrypted in another system"? Surely if a set of
objects A *really* contains pornography, there are certain characteristics
one would expect of the decryption algorithm? If that algorithm is clearly
arbitrary and supplies a significant portion of the information then it is
suspect (compare a genuinely encrypted large image, which might require only
a 128-bit key). Saying that the Bible "contains" any other possible pattern
isn't fair if to prove this you have to supply a "key" that only works for
that one object and contains a comparable amount of information to the
target - that would be like saying the single bit "1" contains the entire
decimal number 12345678 because you can decrypt it using the key 12345677
and the algorithm "add the key to the bit".

Does this matter for what you're saying?

I have a feeling that there is something special about the *properties* of
"real" objects/patterns. An image has the property of spatial contiguity,
for instance, which is not present in an encrypted form of it. There will
presumably (as you'd no doubt say) be a cryptographic universe in which a
JPEGged image does have spatial contiguity because space itself is precisely
as convoluted as the rearrangement of pixels caused by the JPEG compression.
But in such a universe lots of other things would fail to work. A
non-encoded object couldn't move smoothly through space because the space
isn't Cartesian, for instance. And I suspect that JPEGging that object
wouldn't work because the JPEG algorithm is context-dependent. But I can
concede that there could be a universe in which your chair would behave like
the Mona Lisa does in our world.

The phenomena that I'm personally referring to as "patterns" DO have special
properties and aren't arbitrary. Maybe "patterns" is a bad choice of words.
The crucial factor is that they are self-maintaining in some way, so that
once formed they hang around. Atoms resonate, creatures metabolize and
reproduce, memes propagate, societies legislate. Only a tiny fraction of all
possible patterns have this property in any given universe. I guess there
may be a design for a universe in which something that to us looks arbitrary
is capable of self-maintenance, but that in itself doesn't suggest that such
a universe exists anywhere.

I'm still struggling here to see what you're saying, I think. I don't feel
comfortable with the idea that all patterns or interpretations are created
equal. My whole life has been spent trying to find and understand the
special patterns - the persistent ones - or rather the underlying principles
that arise time and time again in emergent phenomena. It's hard enough if I
only have ONE universe to search...

> I don't feel
> comfortable with the idea that all patterns or
> interpretations are created
> equal.

I can help with that. If this view is correct all patterns are not
created equal. Some patterns require particularly contrived
interpretations (or mappings as you put it) to reveal them. A small
proportion of all possible mappings will generate those patterns,
relative to the much larger proportion of all possible mappings that
generate patterns that can be produced by less contrived means.

As a simple example:

Getting stories about Jesus out of the Bible can be done with a much
simpler mapping (one described by a smaller algorithm) than the
contrived mapping needed to extract pornography. Pornography would take
a longer algorithm. Many mappings will gnerate stories of Jesus and many
mappings will generate pornography - but the nature of the Bible
prejudices things in favour of stories about Jesus. The proportion of
mappings that produce stories about Jesus will be much greater than
those which generate pronographics images.

So there is a statisical difference between different patterns and
mappings. Some have more "measure" in the multiverse.

This would seem to be mean that, in principle, your entire thought
processes could be caused by some weird mapping that generates your
mind, and just your mind, with fake perceptions that just happen to have
made sense so far, from the arrangement of matter in someone's desk.
However, the mapping needed to do this would be so hugely contrived that
any instances of mappings like that would be vastly outnumbered by
mappings that generate your mental states from a brain that is set up to
have those mental states, which exists due to atoms being set up, etc
etc.

This is extremely subject to statistical analysis in my view and means
we should expect the world to act fairly sensibly. Yes, there may be an
infinite amount of Steve Grands, all generated by different mappings of
different objects, but any Steve Grand should expect the Sun to rise
fairly normally tomorrow because he should think it likely that he is
one of the "conventional" ones, that represent the vast majority of
Steve Grands, that exist due to fairly conventional mappings.

> porn you would need a different decryption algorithm (or key)
> in every
> single case. And you would have to try a huge number of keys
> to find one
> that worked. Moreover, those keys would be roughly as complex
> as the thing
> they claim to decode. Isn't that a "practical limit on what
> they meant by
> saying one thing was stored encrypted in another system"?

I would say that the practical limitations of the thought police cannot
have anything to do with what exists or does not exist. Do the thought
police control what is "real" by their ability to decrypt stuff? If it
is possible, in principle, to decrypt something and extract a pattern
from it, we should view that pattern as real or not without considering
what the thought police know or do not know, what they can and can't do
and what keys thay have or don't have. In reality, yes, it would be a
practical limit, but how a state's police and courts function in
practice should not concern us as they have no authority to control what
is real - in theory they could just arrest you and say, "Look - we don't
have a key to decrypt it, but we can be sure, having read Almond's
articles, that such a key must exist to extract the porn from your bible
so we are arresting you."

In practice, by the way, I am sure the police could generate some
trivial mapping to extract the porn. In the worse case they could simply
stick lots of randomly selected letters together to make a pornographic
story and simply say the mapping is selecting those letters in that
specific order - or using a different example, you could take any image
and extract any other image you wanted using an appropriate one-time pad
and applying it to the bits. Of course, getting any "clever" mappings
would be harder.

> target - that would be like saying the single bit "1"
> contains the entire
> decimal number 12345678 because you can decrypt it using the
> key 12345677
> and the algorithm "add the key to the bit".

Yes. This is not really about "encryption" - which I am just using as an
example, but that is what I am saying. You can take the digit "1", apply
some algorithm to it and generate the sequence you mentioned. This is
the only way we ever find that patterns exist and either we admit that
all such patterns correspond to real objects, with the very significant
qualification of measure that I mentioned, or we have to justify where
we have the cutoff point.

I would also say that I don't know if I like the word "contains" - it
implies some kind of spatial nature to this, and I view "space" as just
a particular emergent property of our local universe like everything
else. I would prefer to say "imples" rather than "contains" - any object
implies any other object with varying degrees of measure due to various
mappings. The atoms in my desk imply the desk - they don't contain it.
But yes - the basic idea of what you said there is right.

> once formed they hang around. Atoms resonate, creatures
> metabolize and
> reproduce, memes propagate, societies legislate. Only a tiny
> fraction of all
> possible patterns have this property in any given universe. I
> guess there

As you said in your comments on Searle - you could produce some
patterns, using a very contrived interpretation that does all these
things internally, on its own terms. You could find some object, using a
contrived interpretation, that seems to be resonating quite happily if
you are able to keep looking at it and making complex interpretations. I
would actually say that our time is also merely part of "our pattern"
and is not important anyway - time would not feature in many of these
objects and would just be another emergent property. I don't think
space, time and causality are very important unless you live in an
object that has them and need them to get through life.

> equal. My whole life has been spent trying to find and understand the
> special patterns - the persistent ones - or rather the
> underlying principles
> that arise time and time again in emergent phenomena. It's
> hard enough if I
> only have ONE universe to search...

And I still say that approach is justifed - the patterns you see are
special - they are the high measure ones, relative to the measure of our
own universe, that are accessible to your brain. Your own brain limits
what is real to you because its limited processing power makes some
patterns and relationships seem "real" because they are more accessible
to it. Sure, you can augment your brain with computers and advanced
maths to find deeper patterns, but this neurological limitation is
likely to prejduice is in thinking that there is some "cutoff" point.

Another way of looking at things is this. Suppose you were an alien with
1E+12 times as much processing power in your brain. You could be having
exactly the same discussion with me now, but the patterns you were
finding to be the "real" ones could be much more complex, yet they would
seem to you easily accessible and trivially real. To you or I such
patterns would seem horribly contrived - you would need a hige algorithm
to "find" them and that would be well beyond the range of a human brain.

> But in such a universe lots of other things would fail to work. A
> non-encoded object couldn't move smoothly through space
> because the space
> isn't Cartesian, for instance. And I suspect that JPEGging
> that object
> wouldn't work because the JPEG algorithm is
> context-dependent. But I can

They would only fail to work to someone who expects them to interact in
simple, everyday ways, with the things you experience in your space. The
space may appear cartesian to scrambled up observers inside it, sitting
on scrambled up chairs and employing scrambled up JPEG compression. I am
actually suggesting that the space and time you exist in "now" is no
more real than this and is simply the result of some such interpretation
of something else - probably one of the simpler interpretations that can
be made. In fact, your universe could be said to be the messed up one
from their point of view. Of course, if their universe is derived from
yours by a mapping - then it does have some special status which they
should be grateful about, but they would never be able to see this of
course.

> I don't feel
> comfortable with the idea that all patterns or
> interpretations are created
> equal.

Which at least means you get the point about how seriously "out there"
this issue is. Anyway, as I said - they aren't, due to statistical
differences. Maybe that makes things easier?

To close, I suggest we formalize "finding objects/patterns/emergent
properties" and see if we can find common grounds. I propose an
"emergent property detector" which has probes which can take any
measurements it needs in reality. It can move the probes around on
ghostly tentacles. (This thing is a bit like the FSM!) and it employs an
algorithm to control the probes, take the measurements, analyse the data
and produce some "pattern" or higher level representation that it has
found.

E.g. one of these machines, used to analyze the Mona Lisa, with a very
simple algorithm, could easily produce a bitmap of the Mona Lisa. With a
slightly more complex algorithm it could analyze a weakly encrypted
bitmap of the painting and produce the same. With some algorithms it
analyzes your brain and produce a "mental state" - some kind of
description of what it means to be "you" - your mind.

I say that any existence of any pattern can be represented by such a
system working and finding it and that the only issue then is what
algorithms (or mappings as you put it) do we permit and what do we ban
when it comes to deciding which of these patterns have the ontological
status of "existence". If you want a cutoff point when the algorithms
are getting stupdidly long, why? If not - you are in a multiverse. If
you want you can reject the very idea of this machine as a way of
describing our involvement with patterns, but that seems to be
suggesting that something informal is going on.

If this is right, by the way, even things that seem high measure to us,
like tables, chairs and our own minds, may not have any really high
measure - because our whole universe may be a very low measure object in
comparison to someone else's. Logica suggests however, that it is more
likely that universes generally like ours are typical and that our
situation is common.

If you really want fun with this, if it were right, what would it mean
if you were to be suddenly exterminated, given that in such a model the
existence of numerous other versions of you that would carry on with
similar memories, would be a given? I lack any answer to that one.

One final comment - religion and Searle have absolutely no place if this
is right. With regard to religion, you used the great example of
analyzing the number "1" but do you even need the number "1"? What if
you applied interpretations to nothing at all? No input for the
"emergent property detection machine"? If we are being consistent we
should accept that as well - yet nothing is required to actually do this
computing - things exist, in this view, purely by the logical
possibility of doing that computing, in the same way that 2 eggs and 2
eggs make 4 eggs due to the logical possibility of counting them. This
means that the question "Why is there anything?" is meaningless -
suppose we had absolutely nothing - even that implies mappings of it
which produce an infinity of objects, some of which contain descriptions
of time and space and give rise to other objects etc etc. A view like
this makes everything mandatory without having to assume any physical
world to being with.

and I suggest this is where your own views about the ontological status
of "abstract" and "physical" things leads.

I want to address this important comment by Steve in more detail:

> supply a "key" that only works for that one object and contains a
> comparable amount of information to the target - that would be like
> saying the single bit "1" contains the entire decimal number
> 12345678 because you can decrypt it using the key 12345677 and the
> algorithm "add the key to the bit".

This is a natural thought - and it would be strange if this objection
was not made. The idea here is that if you need more information to
"find" something than is in it you are not really finding anything.

To put this another way:

Any pattern, emergent property, object, etc (I think Steve and I give
all this kind of stuff the same ontological status of "things") is
"found" using an algorithm - let's call it the "finding algorithm". Some
algorithm, applied to the atoms in my office could "find" the room
pressure distribution, my furniture, etc. A finding algorithm applied to
my computer could find my computer. We can consider various kinds of
finding algorithm. The best way of imagining a finding algorithm is
inside some machine with lots of probes that it can move around under
control of the finding algorithm, which analyzes readings from the
probes and generates the result. The result could be a 1/0 for if it
finds that property/pattern/object or a more sophisticated version could
actually produce some abstraction corresponding to that property. E.g.
you set the machine up next to an electronic computer, let it "sniff"
around inside the computer with its probes and a program listing appears
on the screen of the machine, indicating that it has found that program.
Different algorithms may detect room temperature, "solid" objects,
people, software, mathematical shapes, fractals etc

The first thing I want to see if we have common ground on is this: for
any object we could say to exist there is certainly such a finding
algorithm. As evidence for this I suggest that any process the human
brain goes through to detect a "thing" must correspond to this in some
way - with information about the environment going through some
algorithmic process in the brain to "find it". It should therefore be
obvious that all objects have corresponding finding algorithms.

Steve and I seem to agree that "existence" is nothing to do with any
disagreement between "physical" and "abstract" - we regard it all as
ontologically the same. Therefore, any different between me and Steve
should be about what permitted "finding algorithms" are.

Steve has made the excellent point that contrived finding algorithms may
be much larger than the things they find.

E.g. I use a 1 trillion bit long finding algorithm, in this finding
machine, to "find" a 1 million bit bitmap image "hidden" in my desk.
"No!" says Steve, "You are just playing games here. The finding
algorithm is much larger than the thing it found - the fact that it is
larger at all clearly means it is finding something that does not
exist."

I reject this by asking a simple question: if there is a maximum
permitted size L for the finding algorithm to find some object with B
bits in its abstract description (e.g. an image of B bits), what is the
cutoff? The obvious answer is to say when L=B but I think this is very
suspect. Suppose a human consciousness were downloaded into a computer
in a simple way - Steve and I would accept it exists. Importantly, for
any object L will always have some minimum number. You can't just say
L=0 is the only valid finding algorithm size - even the human brain
needs some value of L > 0 just to find a brick. What if L is made larger
- the mind is "hidden" in some clever way. I suppose Steve may say it
probably exists. What if L=B? Would Steve say this is invalid? What if
L=B-1? B-2? B-3? What is special about this magical number "B"? Why does
something profound happen when L hits B? If someone was hiding inside a
computer and hid himself in gradually more sophisticated ways,
increasing his value of L (possibly to evade pursuers for example), are
we really supposed to think that the version of him that is in there
when L=B-1 corresponds to a happily existing, hidden AI system, and yet
when L=B he suddenly ceases to exist?A difference of ONE BIT in the size
of the algorithm needed to find him could give him the onbtological
status of a non-existent person? I find this untenable and I think it
casts serious doubt on the idea that an object is somehow unreal just
because the finding algorithm is longer than the object's description.
If we don't like this how are we supposed to deal with it?

For me the problem is simple: I don't deal with it. I don't need to. I
just accept all values of L,even if they are much greater than B. For
Steve the problem is going to be declaring some cut-off point for L
which corresponds in some hard way to the endpoint beyond which patterns
are "made up" and do not correspond to things that exist. One way of
trying to get round this would be to suggest that existence gradually
fades away as L increases - but doing that is only going to raise the
question of what you mean by "fade away" anyway. In fact, accepting any
value of L implies this anyway, as aobjects which can be found by
smaller algorithms will be found by more algorithms, giving them a
greater representation in the set of all objects. Any idea of "fading
away" pretty much invites a multiverse in some sense if you try to
define what this "faded" existence is supposed to be like, because it
leaves all things with some "measure" of existence: they won't go away.

Steve's objection here really is at the centre of this issue. How big
can L be is THE issue here and what someone thinks about this determines
what they think reality is like. My own answer -L has to be less than
infinity and no issues then arise. I view this as the natural
end-product of any attempt to remove the distinction between "physical"
and "abstract" - the value of L is what should next be considered and
the restrictions we place on it everyday, without thinking usually, have
to go, no matter how uncomfortable that is.

So, Steve, that is my issue. For an object containing B bits to have the
ontological status of existing, L < ?. Why? What is so different about
going over this value by one bit that causes objects to cease to exist
and uploaded minds to, presumably, meet certain death if they hide
themselves too well?

> Steve and I seem to agree that "existence" is nothing to do with any
disagreement between "physical" and "abstract"

Well actually I'm not sure if we do. I would say that some patterns, such as a mind, are neither more nor less concrete than some others, for example molecules. On that basis a mind is NOT an abstract pattern. I think it's a real thing, and I think both it and molecules are "patterns" (for desperate want of a better word) of equal ontological status. But that doesn't mean I think ALL patterns are real things. Some patterns that other people might regard as abstract I regard as concrete (or as concrete as anything can be). But I'd say there are also abstract patterns (such as the Mona Lisa or the Theory of Gravitation) that are just that: abstractions. They exist only in our minds. This seems to be our point of disagreement.

> The first thing I want to see if we have common ground on is this: for
any object we could say to exist there is certainly such a finding
algorithm. As evidence for this I suggest that any process the human
brain goes through to detect a "thing" must correspond to this in some
way - with information about the environment going through some
algorithmic process in the brain to "find it". It should therefore be
obvious that all objects have corresponding finding algorithms.

That's interesting. Does the brain operate a finding algorithm? I suppose you could say so. But is it just some *arbitrary* finding algorithm? That I'm not so happy with.

Does each object require its own finding algorithm in the brain? No, that's the fundamental principle of intelligence, in a way - its generality.

Could someone else's brain look at what I interpret to be your chair, and see Albert Einstein? In principle, yes, but what kind of screwed-up childhood would such a person have had?!? Is there a universe in which such a brain exists? If so then it would have to be a universe in which what I perceive here as a chair not only appears to be Albert Einstein to this alter-ego but also *behaves* like Albert Einstein (this is actually redundant, since that's what "appears to be" means). Whatever finding algorithm this person's brain implements must also be implemented by spacetime in general, for there to be any self-consistency in this other universe. What behaves as a chair in my universe must behave entirely like Albert Einstein in his, whether he's there to witness it or not. So isn't this other object just Albert Einstein after all?

I think I'm probably getting confused again...

But in response to your specific question about cut-off points, I would certainly suggest that something with an information content smaller than the thing you want to "find" in it doesn't count, thus setting a lower limit. An atom can't be said to contain Neils Bohr (to switch physicists for a moment) because it simply doesn't contain enough bits. To claim that it could represent Neils Bohr would be like an art critic writing a long essay on the deep artistic meaning of a pile of bricks - the meaning isn't in the pile but in the critic (or more charitably towards art critics, in the artist).

This business of how many bits something contains (not that I'm a fan of overstretching information theory like this) raises the question of what delineates a pattern in the first place:

I'm probably going to regret saying this, because I'm a big believer in what Douglas Adams called "the interconnectedness of all things" - everything's a messy blur to me and we draw boundaries around things at our peril. Nevertheless, to me, an atom is a bona-fide discrete thing - an object in its own right - because it consists of those relationships that wrap around on themselves and are necessary for it to preserve itself over time, and it corresponds to the arrangement that actually does persist over time. Any interactions beyond this self-maintaining "set of vibration modes" are better counted as *properties* of the atom, not components of it - they're the channels through which the atom interacts with other atoms and through which complexes of atoms give rise to higher emergent phenomena such as enzymes.

Similarly a human being doesn't extend beyond its body (although it is the arrangement of its atoms that persists, not the atoms themselves, this persistence is localized and circumscribed). To take something less tangible, my mind is a discrete thing - it interacts with your mind but we're separate entities in a meaningful sense - the boundary between us isn't just arbitrary. I am not you.

But if all abstract patterns are accepted as real, where are the boundaries between them? Is the Mona Lisa a thing or is "the thing" the Mona Lisa plus its frame? Or should we include the Louvre? Or its visitors? Or is there only one pattern - this entire universe and its infinite possible interpretations? Where do you draw the lines?

One thing I want to clarify is that I don't think a chair, for example, "contains everything as that implies some kind of spatial relationship. Rather, I would say it implies everything. The relationship is one of implication.

Steve said: "would say that some patterns, such as a mind, are neither more nor less concrete than some others, for example molecules. On that basis a mind is NOT an abstract pattern. I think it's a real thing, and I think both it and molecules are "patterns" (for desperate want of a better word) of equal ontological status. But that doesn't mean I think ALL patterns are real things."

This is really the point where we differ. I am interested in the "cut-off point" here. Any philosophy which admits some abstractions as real objects, as emergent properties, and not others has the problem of needing some definition of a cut-off and a justification for how it occurs and why - a bit like the need to find a wavefunction collapse mechanism in quantum mechanics, as a very loose analogy. Your position needs some kind of "ontological collapse mechanism" - and my view is simply to abandon any attempt to get one.

Steve said: "That's interesting. Does the brain operate a finding algorithm? I suppose you could say so. But is it just some *arbitrary* finding algorithm? That I'm not so happy with."

I do not think it is arbitrary. It is developed to find those patterns, those emergent properties which are relevant for survival. In particular, those which fit into the world at its own level. These will tend to be a subset of those which can be extracted using finding algorithms that are not contrived. I know there is a table in front of me because my brain can extract that emergent property from the data using a relatively uncontrived algorithm. My brain is not going to get involved in extracting stuff which may as well not be here for all the effect it has on us - things which would take longer algorithms to extract.

Steve said: "Does each object require its own finding algorithm in the brain? No, that's the fundamental principle of intelligence, in a way - its generality."

I agree on this. We both have an interest in how such an algorithm may be made: I think we have both tried to make one. The brain clearly uses some process equivalent to a general algorithm, or a set of general algorithms. However, any object which can be found by a general algorithm can surely be found by a specific algorithm for that object. The general algorithm is just used for efficiency. It makes the problem tractable. The general algorithm allows a process equivalent to simultaneous running of all the specific finding algorithms that do not now need to be explicitly run. I think it is obvious that for any object a human brain can detect, a specific algorithm exists that would extract that pattern and "find" it, and that this is the only reason the object can be found by a general algorithm. I further suggest that this is the only evidence we have that any objects exist. You know a football exists because the general algorithm "finds" it. The general algorithm can find it only because it is findable by a specific algorithm. Therefore, all the neurological processes of seeing a football are telling you are that it can be found by some algorithm - yet you find that sufficient for declaring a football to exist. Of course, the finding algorithms in the brain are not arbitrary. Unless you are insane, as you point out, you will be using a very restricted set of economical finding algorithms, with very limited amounts of data in them, that find things that closely relate to all the other things you can economically find. Yet that does not mean these "economically found" objects have a special ontological status - merely that they are easier to find and that evolution favours that approach in your brain.

Would you at least accept that much - and if you do, where is the cut-off? Is there anything special about the amount of data in the finding algorithm matching the amount of data in the found object. I still suggest there is some arbitrary cut-off point in your ontology - so arbitrary you are going to have a lot of trouble stating where it is. Even if you base this on how something relates to other things that is circular. You may say X, Y and Z exist, and B is real because it relates in some "solid" way to X, Y and Z, but that does not say why we can't say that C exists because it relates to E, F, G and interacts with them just as deeply. You have to arbitrarily start declaring some objects as the "grounded" ones to get going here. Further, it does not give you a decisive cut-off point. I can see easily that some other universe, with pixies in it, implied by some extreme finding algorithm applied to some bit of our universe will appear "divorced" in some way from everyday life. The pixies won't steal our stuff. We can't talk to them. We can't trade with them. They don't bump into us. The problem is, you don't just jump from everyday things that we can bump into to pixies. As we make the finding algorithms gradually longer, and they become more extreme, the objects found will seem decreasingly interconnected in any meaningful way with everyday life - but there is no cut-off point at which this starts - it is a gradual process as L increases - so if you are going to base your ontology on eliminating, from physical existence as "real things" all those entities found by algorithms that don't interact with our everyday stuff, where do you draw the line? How low does the interaction go before you say, "Enough!” and declare that this thing is made up?

Steve said: "But if all abstract patterns are accepted as real, where are the boundaries between them? Is the Mona Lisa a thing or is "the thing" the Mona Lisa plus its frame? Or should we include the Louvre? Or its visitors? Or is there only one pattern - this entire universe and its infinite possible interpretations? Where do you draw the lines?"

All of these things would be regarded as real in their own right. You could have some real object A and a real object B containing A, and a real object C containing A, B and so on. I would argue that your mental state, your feeling of being "you" is best considered as an emergent property abstracted from the underlying "hardware" (and I know neither us thinks that word formally means anything) in the brain. What this means is that you, and the things you are aware of around you are a kind of special case of things that are ontologically "close" due to the specifics of the extraction. Anyway, you have this problem of boundaries, if you think it is a problem, even if you accept only a subset of patterns.

An important point is that of measure. Just because objects A and B exist does not mean they are the same. A will be represented more in the set of all objects if A can be found by a smaller algorithm - because that will mean that a bigger proportion of the set of all finding algorithms can find it.

I want to make it clear as well that I only use decryption as an example. In reality, I think any object that can be said to exist has some corresponding "finding" algorithm and those objects we see in everyday life are very accessible to us by being very easily extractable from the underlying reality.

My basic point stands: Where is the cut-off? or do you think we don't need a cut-off?

I want to also suggest that the problem of splitting or combining computers containing minds causes big probability problems if you try to deal with them outside this kind of ontology. That is dealt with in my articles on this:

http://www.paul-almond.com/Substrate1.pdf
http://www.paul-almond.com/Substrate2.pdf
http://www.paul-almond.com/Substrate3.pdf

Although these articles do not propose that a multiverse has to exist, they would have to if continued further. Otherwise it just gets impossible to consider probabilities of being in different situations in the sorts of thought experiments considered in those articles. I would be happy to introduce one of these thought experiments into this discussion if you wish, so that we need not go to external articles which are long.

> This is really the point where we differ. I am interested in the
> "cut-off point" here. Any philosophy which admits some abstractions as
> real objects, as emergent properties, and not others has the problem of
> needing some definition of a cut-off and a justification for how it
> occurs and why

That's a fair point. I still dislike the word "abstractions" but you're right that I need to come up with a justification for my choices. It needn't necessarily be a cut-off point though - it can be qualitative instead of quantitative, surely?

So, what is an emergent phenomenon? I'm not aware of any definitive explanation elsewhere, so I'd better come up with something that suits my needs.

Firstly I'd say it has to be a comparatively common occurrence. A phenomenon isn't emergent (it isn't even a phenomenon) if it is unique. It must occur many times in different places and thus be worthy of a categorical name.

For our purposes it must also be persistent over time and resistant to dissolution - a splash isn't an emergent phenomenon but a hurricane is.

These two facts will be satisfied because, it seems to me, an emergent phenomenon always embodies some kind of attractor. Say you're rowing across a glacial lake in Canada (because that's where I was when I first really thought about this) and you splash the water with your oar. Large amounts of water will be churned up into all sorts of shapes. Before you know it, though, the noisy surface will have subsided and most of these forms will have disappeared. From the melee will have EMERGED some examples of the few stable forms that can exist on water: vortices and ripples being the most common. Any initial vortex-like shapes will become more vortex-like over time, and a vortex disturbed by an external force will tend to recover, because vortices represent an attractor in water space.

Emergent phenomena also involve the transfer of something in order to maintain form - either feedforward or feedback. A ripple persists because it propagates its shape (by the transfer of energy) forward in space. A vortex is more sophisticated because it wraps around on itself, maintaining its form without having to keep running forward. A simple life-form consumes energy to reduce local entropy, using feedback to maintain homeostasis.

An emergent phenomenon is never, therefore, a static timeless pattern - it's a dynamic process. Despite all this talk of chairs and suchlike, I don't really think a chair is an emergent phenomenon except in the form of a meme - a design that is propagated from one person to another and maintains itself through popularity. The Mona Lisa is a class of meme called a portrait. Memes are very weak forms of object. Societies are stronger forms, despite being spatially nebulous, because they maintain themselves through kinship and laws, and retain their form against immigration and invasion.

Many emergent phenomena CREATE static, relatively timeless patterns - a hurricane is spiral-shaped from above, for example. Intelligent beings sometimes recognize other emergent phenomena by their spatial arrangement, but sometimes their senses detect other properties, such as the way they respond to a stimulus or the noise they make. The first fact - that emergent phenomena recur frequently - gives us memories with which to recognize them, and experiences with which to categorise them and predict their behavior. This is the foundation for intelligence.

The universe contains "affordances" that make certain emergent phenomena (certain attractors in state space) possible and others impossible. But the existence of a new class of emergent phenomena creates new affordances, making other forms suddenly possible (the universe couldn't discover minds until it had created brains, and these required the discovery of catalytic networks).

Each layer of phenomena makes new properties possible, and given a certain set of properties there will likely be various arrangements of relationships between the properties of components that in turn make other classes of phenomena possible (just like the properties of transistors and capacitors arranged in the right circuit can give rise to oscillators - a new level of being). The universe is constantly bootstrapping itself up into new levels of being.

That's about the gist of it, for me. Some patterns are special because they fit these requirements for persistence - whenever they come into existence they stay. And their existence makes other forms possible. Some are very special - molecules can make a huge variety of new forms of existence, so can computers; life is amazingly persistent compared to less "intelligent" patterns of interaction.

So I don't think it's a question of a quantitative cut-off point on a continuum. I think it's a qualitative thing to do with properties (i.e. ways of forming relationships).

Does that mean we're actually talking about different things?

- Steve

It seems we may have different semantics about what an "emergent property" is, and that may cause us semantic problems. I would regard a chair as an emergent property of the atoms underlying it. An individual atom by itself does not have the property of "chairness", but many atoms in the aggregate do have the property of "chairness". It can only meaningfully make sense to say they have this property because they satisfy some arrangement, some pattern, and the obvious way of generally describing patterns is with algorithms - specifically Universal Turing Machine tapes (but you knew I was thinking that anyway - I am just saying that for the benefit of anyone else). You seem to have some more specific requirements for something to be an "emergent property", so maybe we could discuss what is "real"? Out of all these patterns that could be "found" by "finding algorithms" on this continuum from short finding algorithms to long finding algorithms, you clearly reject most of them and I accept all of them. You are suggesting that your reason for acceptance or rejection of a particular object is not due to where its finding algorithm is on this continuum but due to other properties - and you have described those properties in your last posting. Would that be a fair description?

If so, I would make the following points:

1. Can you justify this as criteria for what is real? Is there any philosophical argument to support the idea that these things are real and other things are not real? I could just try to declare your criteria arbitrary with regard to existence of things. I could say that, rather, they describe things of which you are going to be particularly aware.

2. It seems to me that any of these "qualitative properties" you mention could be present in varying degrees, and that the transition from something that has one of these properties in abundance to something that, for all practical purposes, lacks it, is not abrupt but instead we could imagine lots of things in between that have whatever property we are using in varying degrees. Even if we adopted one or more of your "qualitative properties" as the criteria for something being real would that not just give us another continuum to worry about?

3. With regard to one of your points about an emergent property never being timeless, I should point out that, the way I see things, space and time are not particularly important. You could make lots of patterns involving space, lots of patterns involving time, lots of patterns involving both, and lots of patterns for things derived from such patterns in fairly simple ways so that the are effectively objects in space-time. In other words, I would suspect that what we call "space-time" has no fundamental nature and is merely a feature of the object that provides the physics for our universe, and from which we are derived in a comparatively simple way. For that reason, I am not going to try to define ontologically profound properties in terms of space or time - the same for energy really. I regard the entire framework of physics, space and time as merely a local phenomenon. This in itself is going to cause a gap between us in how we think of things. Everything else humans have ever thought of as "the framework" has been found out to be strictly local - and I think it will the same story with space and time. I think that we are still, generally, hung up on some notion of a "fundamental framework" of space and time, when the only reality of the universe is logical relationships.

4. This will make it obvious I hold out no hope for a "theory of everything", beyond general statements about the logic of how objects imply other objects: all physics is local physics.

5. If you need some properties like persistence over time, particular behaviour, etc, you could still get these with extreme interpretations. For example, a very extreme interpretation could pull something implausible out of reality that does seem stable over time, and that does many of the things you require - but simply does them while being an ultra-extreme interpretation. For example, I am sure some interpretation, in theory, allows me to pull the film of Casablanca out of my desk in real-time and play it in real-time: you need not do the pattern extraction at a single moment. If you can use extreme interpretations you pretty much get anything you want so I would suggest the limitations you place introduce have the cost of introducing some arbitrariness while lacking the benefit of avoiding all the extreme craziness. Even if we needed properties like persistence over time, I would argue we should ask "whose time?" - something could persist in a time described by some pattern pulled out of an instant of our universe, on its own terms.

Two questions, which probably are a bit tactical in nature:

1. Do you think my view is inconsistent with what we observe?
2. Whose view is more complex? Mine or yours?

Ah, yes, it looks like we're using the same words in different ways. Isn't that always the problem?!?

Am I playing Samuel Johnson to your Bishop Berkeley at all, maybe, perhaps? I seem to detect a degree of Idealism.

I can see why you use the term "emergent property" when describing a chair. None of the atoms or parts of a chair can be sat on, so chairness must be an emergent property of the whole. But I've been talking about "emergent phenomena", by which I'm referring to the entity itself - the whole panoply of properties - and I'm using it in a very much less anthropocentric way.

A property, I imagine you'll agree, describes the nature of a relationship. Things don't just "have" properties, independently of anything else. Mass describes the force exerted between a given object and another, or the force needed to accelerate the object in a particular reference frame. Size is a comparison with an external standard length. Wetness describes the way that something wet behaves when it wets something. As far as I can see, all properties describe how something relates to something else.

Some properties only have meaning relative to human beings. Reflectance is a property that shows how an object interacts with photons, but color is an invention of the brain - objects don't "have" color. Chairness, obviously, is a property of this kind - it means nothing whatsoever without someone to sit on it. If the human race were wiped out tomorrow, all chairs would instantly cease to be chairs and become meaningless clusters of wood and metal.

All properties exist in terms of relationships, but the contribution of each party can vary. In the case of the kinds of phenomena that interest me, the properties are largely contributed by the phenomenon itself. Moreover, the phenomenon as a whole - its panoply of properties; its ability to enter into relationships and create interactions - has meaning independently of a human observer. Tornadoes would still exist, even if there were no Dorothy and Toto to be whisked up into them, but chairs would cease to exist, because chairness is a property that is ascribed to chairs (you can sit on a rock, but a rock is not a chair unless you allow it to be).

For me, i suppose something can only be real if its chief properties remain when there's no-one there to look at it. (Societies are made from people, of course, and minds are people, but in both cases there's no need for an external critic to ascribe properties to them; societies and minds have effects on the inanimate world, whereas chairs and paintings don't). I think that's a reasonable requirement for what I call real - after all, reality existed before humans evolved and will exist after we're gone. You clearly define "real" in a different way. Does one have more utility than the other? Should we choose separate terms for each?

There are some interesting gray areas. A computer chip and an identical chip in which half the circuit has been printed the wrong way round have massively different properties - one computes and the other just sits there. They have exactly the same information content, mass, constituents, etc. There is no quantitative difference between them, only a qualitative one. So is computation something that we humans arbitrarily ascribe to one chip and not the other? I don't think so but it's tricky.

Anyway, to try to answer your questions:

1. Can you justify this as criteria for what is real? Is there any philosophical argument to support the idea that these things are real and other things are not real?

It's purely an assertion but basically I suppose I reject what I perceive as your more Idealistic view in favor of Realism - hence the name. If you doodle at random, most doodles will look like nothing but some will look like faces. Are those doodles special in any measurable way? I don't think so. I think the term reality should be reserved for systems with coherence, self-maintenance, and "meaning" within the universe at large, not just this tiny corner of humanity.

2. Even if we adopted one or more of your "qualitative properties" as the criteria for something being real would that not just give us another continuum to worry about?

No. It's not the same. The transition from something that has coherence and self-maintenance to something that doesn't is both continuous AND abrupt, in much the same way that a switch transitions from off to on by passing through all shades of resistance in between but without spending long on the journey. It's a quantum transition. Systems may be better or worse than each other at maintaining their form against insult (that's why high intelligence is better than low, for instance) but on an individual basis, once they start to lose it they rapidly go all the way. A person who is dying goes through the stage of being half dead, but they don't stay that way for long. Emergent phenomena (in my definition) represent attractors in state space, and transitions between attractors are continuous but abrupt. They are islands with seas between where nothing exists for long.

3. I would suspect that what we call "space-time" has no fundamental nature and is merely a feature of the object that provides the physics for our universe

I agree with you that space and time are not fundamental constructs and they themselves are emergent properties - neither could exist until distributed matter existed. But I'm not very interested in the first picosecond after the Big Bang, and from that point on "persistence" had a meaning. If atoms didn't persist there wouldn't be any. Staying around for a while is a prerequisite if you want to be described as a real thing in any meaningful sense.

4. This will make it obvious I hold out no hope for a "theory of everything", beyond general statements about the logic of how objects imply other objects: all physics is local physics.

I hold no hope for it either, but to me that's part of the joy of it. The fact that each level of being introduces the potential for properties and interactions that didn't exist before and thus enables a new level of being (atoms begat organisms, brains begat consciousness) means there's no known limit to how inventive the universe might become. We can still identify general principles, and that's very interesting, but we can't predict what consciousness might ultimately give rise to. If there's ever a god in the sense of a consciousness beyond that of individual orgamisms, it'll be an emergent one - a latecomer to the universe, not its creator.

5. a very extreme interpretation could pull something implausible out of reality that does seem stable over time, and that does many of the things you require - but simply does them while being an ultra-extreme interpretation.

You're back in your Bishop Berkeley robes again! I think I've already stated my view that reality is independent of interpretation by humans. If by interpretation you mean something more automatic, capable of being carried out by an inanimate universe, then we need to discuss this further - it's roughly akin to "measurement" and the issue over the Copenhagen Interpretation of quantum theory.

1. Do you think my view is inconsistent with what we observe?

I think it's logically valid. But logic, like mathematics, can form statements that are grammatical yet have no semantics. I'd place it alongside things such as Zeno's Paradoxes.

2. Whose view is more complex? Mine or yours?

I don't know how to answer that. Complex in which of several possible senses?

EDITED 3/27/09. I just thought of an illustration that might be worth adding here (although it might get us bogged down with quantum theory - I could probably think of non-quantum-level examples but this one's easier to relate):

An atom can be in one of several quantum states - let's call them 1, 5, 7... Think of the states as being different stable vibration patterns. It's easy to set up experiments to determine which of those states a particular atom is in. In other words, your finding algorithm is straightforward.

State number 3 would be a different pattern, intermediate in form between 1 and 5, but try as you might, you'll never find an atom in that state because it isn't tenable - anything that passes instantaneously and immeasurable through that state is, like the throw of a switch, irrevocably committed to ending up in state 1 or 5. I would say that atoms in states 1 and 5 represent reality and, although an atom in state 3 is something we can logically conceive of, it doesn't really exist. But you would presumably claim that atoms in state 3 exist too, it's just that their finding algorithm is more complex?

Yet there's the hint, surely? If a simple observation is needed to interpret the atom as being in any of the states 1, 5, 7, etc. but a hugely more complex algorithm is suddenly needed to "show the existence" of a state 3, doesn't that suggest that state 3 is a figment of our imaginations and not real? Surely 1 and 5 have a radically greater right to be regarded as real, not just because their finding algorithm is simpler in absolute terms but because there's a sudden inexplicable discontinuity between the algorithm needed either side of "finding" 3, and the one needed to find 3.

Does that bother you?

I wonder if physical entropy has something to do with this.
My naive understanding of this term is "having structure" in contrast to being "uniformly random". Of course there is an exact definition too, which I find a bit hard to imagine however.
Entropy also assigns some "value" to information or - in a physical sense - states of energy. It is always possible to convert "valueable" low (!) entropy states into states of higher entropy. It is however not possible to revert this process and decrease entropy of the system, without introducing some additional object/information/state/what ever that has low entropy.

In other words: In order to get porn out of the bible, one has to "consume" a lot of low-entropy information. The same is certainly true to make a desktop look like the Mona Lisa (in this case, the low entropy thing is the extremely "structured" way one has to look at the desktop), while making the Mona Lisa look like Mona Lisa does not require such.

Happily our planet (currently) emits more entropy into the universe than it gets (mostly from the sun). Only this fact makes it possible for life to evolve (or even for molecules or atoms to exist), since life forms generally have a very low entropy and as said before things of low entropy can only emerge by increasing entropy somewhere else. This entropy than has to go somewhere, e.g. by being radiated into outer space. Otherwise all the entropy on earth would constantly increase and finally would not allow any structure to emerge any more.
The whole earth would just be a uniform unstructured mess of energy (this is my interpretation, don't know if physicists agree in this last point or think that it would look somewhat differently).

Just some note at the end: It seems that the term "entropy" is also used for some metric in information theory. I don't know if both the physical and informational meanings are interchangeable. I referred to the physical entropy in this comment and when I was talking about information, I really mean information that is represented by some structure in either the distribution of energy or mass.

PS: Sorry, I did not mean to make this comment appear in the middle of your conversation. Cannot find a way to move it down however...

Steve Grand said: "Chairness, obviously, is a property of this kind - it means nothing whatsoever without someone to sit on it. If the human race were wiped out tomorrow, all chairs would instantly cease to be chairs and become meaningless clusters of wood and metal."

"Chairness" may have been a bad example by me, because it presents these difficulties. I was not thinking so much of a rock someone could sit on (though I could spend a long time arguing round that one maybe and not helping this debate much) as in terms of a typical kind of chair that you can buy at a furniture store. My position would be that if humanity were wiped out tomorrow, all those kinds of chairs would still have some property in common. You could write an algorithm that could be applied to any object and told you whether or not it was that kind of chair (ignoring the more awkward rock case). Even in the absence of humans, that algorithm would still indicate "yes" for chairs (the IKEA kind rather than the rock kind) and "no" for other objects. I do not think that algorithm is reliant on humans at all. If we call that algorithm C we could say that for this class of objects, C outputs "yes", even if everyone is dead. If C were a very long, contrived, algorithm I think my position would be more awkward, but I think for chairness the algorithm would be quite short and economical. Now, I would accept that the semantics of associating C with "chairness" may be more awkward if nobody is around to sit on the things, but I still think that chairs (from places like IKEA, who should pay me for product placement) do have this mathematical fact in common.

Steve Grand said: "It's purely an assertion but basically I suppose I reject what I perceive as your more Idealistic view in favor of Realism"

and I would question whether that assertion serves any philosophical purpose, but you knew I was thinking that anyway. I think we are both likely to regard our own views as "realism" in a tautological way. I think that what I am proposing is "real" is "real", and only that, and so do you. Rather than considering my view idealistic, I consider this the most important feature of it: it demands that any position about what is real and what is not real can be formally described, at least in principle. I think that if any statement is of any philosophical relevance it should be possible to state it formally, at least in principle, and statements about what is and what is not real should certainly fall into that category. I think this requirement of being able to formally state a position is critical. I think my position can be easily formally stated and I think that is less obvious for yours. For example:

Suppose we consider my continuum of "extracted patterns" (I won't call them emergent properties if we differ on that term), each "found" by an algorithm applied to reality. The problem of formally stating which are real and which are not is trivially easy for me: I just say, "All of them." You have a more complicated time. You have to be able to say which are real and which are not, and I think you should be able to produce an algorithm which could do that on your behalf, which could be applied to each of these patterns and tell us whether it is a "real" object, or just some nonsense someone has pulled out of his wallpaper pattern. I think you would find that an intractable task, even in principle. I think that my position can be easily formally stated and yours cannot - and that that gives me an advantage.

Now to deal with your quantum physics example:

Steve Grand said: "Yet there's the hint, surely? If a simple observation is needed to interpret the atom as being in any of the states 1, 5, 7, etc. but a hugely more complex algorithm is needed to "show the existence" of a state 3, doesn't that suggest that state 3 is a figment of our imaginations and not real? Surely 1 and 5 have a radically greater right to be regarded as real, not just because their finding algorithm is simpler in absolute terms but because there's a sudden inexplicable discontinuity between the algorithm needed either side of "finding" 3, and the one needed to find 3."

Why should something have more "right" to be regarded as real just because it is found by a simpler interpretation? I see no support at all for this philosophical rule. From where does this "right" come? I would suggest we can philosophically manage quite well without this principle, that it does not help us make any more sense of the world than we can already, and that we should therefore do without it.

Observations of the 1, 5 and 7 states does not have any special status. Some interpretative algorithm will need to be applied to get them (and it would generally be applied in a human brain). We will be able to find lots of stuff using simpler interpretations, "beneath" the 1,5 and 7 states. From that perspective we could declare the 1, 5 and 7 states to be the results of interpretative algorithms that are too long and contrived to correspond to anything real. We do not do that of course. One reason is that the 1, 5 and 7 states fall into the narrow range of interpretations to which our brains, with their limited computing ability, have access. Another, and possibly better reason, is that these states have a strong relationship and interconnection with all the other stuff of which we are aware in the world. The 1, 5 and 7 states are part of our physics, which relates them, by various means, to everything from cows to fire engines. This may seem to give the 1, 5 and 7 states some special status, but the problem here is that all the other stuff to which they strongly relate is only justified in this way by being strongly related to the other stuff and so on: our view of reality is a set of mutually related and interconnected objects, and any argument that an individual object in this set of "real" things is real because it is interconnected with the others - because it is important to how we understand all the other things and make predictions, etc - ignores the fact that the entire set of objects only contains those particular objects, and not others, due to limitations in how much interpretation our own brains can readily make. A tree seems real because our brains can readily make that kind of interpretation. A cat seems real because our brains can readily make that kind of interpretation. The fact that cats can scratch trees and leave claw marks on them seems to justify cats being real to someone who already thinks that trees are real, and trees being real to someone who already thinks that cats are real. Of course, our ability to form interpretations goes beyond this. We can use formal reasoning, advanced mathematics, pencils and paper and computers to gain access to interpretations that are beyond the "intuitive" abilities of our unaided brains. However, the "secondary", more formal, deeper interpretations that we find significant are still going to be those which have some relevance to the world of cats and trees and any interpretations that we find significant beyond this would still need to be significant to these secondary interpretations, connecting them in some way to the world of cats and dogs. If we need interconnectedness, we could find an entire set of objects, with some extreme interpretations, all of which interconnect with each other (I will say more on this shortly), and the only reason for declaring that invalid would be if I accept your suggestion that somehow the set that comes from less contrived interpretations (i.e. shorter interpretative algorithms) has more "right" to be real - as if we were somehow aiming for metaphysical fairness. This is not just about what human brains are "powerful" enough to detect. A human brain, and the body containing it, exists at a particular level of interpretation, and those things which impinge in any significant way on the mental state of

Earlier, you specified various criteria for things being "real". I am not dismissing those criteria at all. I think you very well describe, in a profound way, the features of those things that have any real, direct interest to us, and that have any noticeable impact on our experience of the world. What I am rejecting is the idea that those criteria dictate what exists and what does not exist and I you have not said anything to persuade me that those criteria have that much ontological power. I still do not see from where such ontological power is supposed to come.

I will, however, partially agree with you about objects with less contrived interpretations having more "right" to be real - sort of, in a very specific way. I will, however, not partially agree with it just as an arbitrary principle from nowhere but for a different reason. Any object will not just be pulled from some other object by a single interpretation. Many interpretations of the same object, or part of an object, could find the same object. For example, many different interpretative algorithms could be applied to my brain to find essentially the same "abstract" (I know neither us likes that word but I will use it so that other readers can understand better) object that we call my mind. To be more technically accurate, provided we can specify algorithms of unlimited length, an infinite number of interpretations can be used to find my mind by making interpretations of my brain. Similarly, you can apply an infinite number of interpretative algorithms to The Mona Lisa to extract the "abstracted painting" or the "treeness" from the atoms of a tree. We should give the status of "real" to every object found in this way - for reasons you will be aware of now, and with which you disagree. Now, we could also apply different interpretations to my desk, or to a single atom in my desk, and extract some object corresponding to my mind. As with my brain, we could apply an infinite number of such interpretations to my desk to find my mind in it. Does this mean that my desk is no different from my brain, in terms of causing my mind to exist, that the atoms in a tree are no different than the atoms in a rock in causing a tree to exist? No, I am not saying that at all. That ignores the very important issue of measure. Suppose we use some machine that looks at all the possible interpretations and we get it to go through all the possible interpretative algorithms with some maximum length of N bits, assuming N is quite large, and starting with the shortest interpretative algorithms first. We first use the machine on the atoms in my brain. It will find an interpretation corresponding to my mind quite quickly, because the arrangement of the atoms in my brain lends itself to that kind of interpretation: it is hardly a contrived interpretation of brain. This is not the end of it, however: it will go on to find more interpretations corresponding to my mind. The same object will be found many times. Suppose we now use the machine on my desk, with the same value of N (the maximum interpretative algorithm length we will also find my mind there, but with a hugely increased length of interpretative algorithm because you need a much more contrived algorithm to pull a mind out of a desk. As we go on, looking at longer interpretative algorithms, we will find more versions of my mind, but we will not find anything like as many as we did when looking at my brain, because of the greater contrivance that is needed to make them. Each of these minds that we find can be considered to be a real mind, but we find a lot more being implied by my brain than by my desk. You could complain that this is only with some finite value of N, but we can say that N tends to infinity and work out the proportion of minds that are implied by my brain and the proportion of minds that are implied by my desk. The proportion of minds, like mine, implied by my brain will be a vastly greater number. This is what I meant when I said that I accepted the idea that the objects which can be found by less contrived interpretations have more right to be real, but in a restricted way: there are a lot more of them. I know you will disagree with this, but I want to point out that it at least gives some recognition to the role that things like brains play in causing minds. The argument is not really as simple as saying that my chair is really Einstein, but rather that my chair can imply Einstein's mind with much less measure than my brain can. Both of us, I think, can easily see that there will the things produced by contrived interpretations in this way will have an insignificant measure compared to those produced by contrived interpretations. I also point out that this not only addresses your concern of "more right to be real" (in a way), it deals with it in a more formalized way than your own observations about contrived interpretations do. All you can do is declare the more contrived interpretations to have "less right to be real". My arguments, in principle, would let us put numbers on that "right" and also gives an ontological explanation of what "less right to be real" means, in terms of reduced measure/lower density of occurrences of that object.

So, to summarize: no it does not worry me, because the idea that I think there is no difference at all between the atom in the sensible states and the atom in the silly state would be a caricature of my position which ignores the statistics of the situation.

Steve Grand said: "I agree with you that space and time are not fundamental constructs and they themselves are emergent properties - neither could exist until distributed matter existed. But I'm not very interested in the first picosecond after"

I think I am taking a more extreme view here of space-time as a construct. If all the things that can be found by interpretation exist, then it would be strange to think of the observable universe that we seem to inhabit as the ground floor of all this. It would be almost certain that the observable universe - everything we know - is just an object that can be found by some interpretation of something else. This means that even our laws of physics, and our space-time would just be mathematical constructs that could be made from some other object by some interpretation. That other object would not need to exist in space or time, and most objects in such a multiverse would not be embedded in space or time. The way in which I view space-time as a construct is as a local phenomena of this object with which we happen to be familiar. Of course, other such objects like our observable universe would exist - giving a multiverse of universes vaguely like ours in terms of having space and time. However, such objects would be in a minority. Words of space and time would be of no relevance for describing most of reality. This is more extreme than saying that space-time would lack meaning without the interactions of the objects embedded in it.

Steve Grand said: "You're back in your Bishop Berkeley robes again!"

This is only about idealism in so much that I think that our position on ontology should be one that can be formally stated. There is nothing wrong in wanting a position that can be formally stated. There is also nothing wrong in a position that does not try to selectively assign the feature of "reality" things without good reason.

Steve Grand said: "If by interpretation you mean something more automatic, capable of being carried out by an inanimate universe, then we need to discuss this further - it's roughly akin to "measurement" and the issue over the Copenhagen Interpretation of quantum theory."

Maybe you think I am suggesting that observers or computation make things real? I am not suggesting at all that some observer, or anything else, needs to run some sort of machine or computational process to make an interpretation to make things real. I think you are suggesting that that would be a bit like the interpretation of quantum mechanics which says that wavefunction collapse occurs (and the probabilities "condense" into real events) when an observation is made. I find the Copenhagen Interpretation hopelessly vague as any workable scientific model, because it does not formally say what an "observation" is - and I doubt that there is even any remote hope of anyone formally saying what an "observation" is. The Copenhagen Interpretation, if we are supposed to believe that is THE description of reality would rely, then, on a process which cannot formally be described - which in my view means it cannot be a formal description of real. I am not saying it is nonsense though: it works well as a kind of informal, practical way of describing reality when we have two different models, neither of which gives the full picture by itself. Some people take the Copenhagen Interpretation further and extend it into the "consciousness collapses the wavefunction" model, which I think is just as bad. Once you take that approach you are into the territory of taking things like "What the Beep Do We Know?" seriously.

I have said this to show that I do not have any belief in ideas that observation makes things real, that consciousness makes things real, or that some sort of "process" needs to be applied to things to make them real. It is, in fact, because I am so opposed to the lack of formalization found in such views - the throwing around of such vague words like "observation" - that I have the view I have about ontology. When I say that a thing exists because a particular interpretation can "find" it, that does not mean that I think something must actually make that interpretation. I do not think that humans make things exist by applying interpretations. I do not think that things exist because some god, or some cosmic interpretation machine is there running through all these interpretations, making things real. Rather, I think that things exist when there is the logical possibility, of applying interpretations to other things to find them. Whether or not anyone or anything actually applies those interpretations is irrelevant. The only time the issue of conscious observers or computers would become relevant would be in the special situations (special to us) when some interpretation can find a conscious mind or a computer.

This does not necessarily mean that I think something is really "caused" to exist by the logical capability to make these interpretations. Whether or not I am saying that would depend on the semantics of "cause". For example, do the atoms in a tree "cause" that tree? If "cause" is meant in the temporal sense of cause and effect the answer is certainly "no". Rather, the atoms in a tree imply a tree in much the same way that 2+2 implies 4. it is not really that 2+2 causes 4, or that 2+2 "makes 4 exist" or that 4 exists "because of 2+2". It is, rather, that whenever 2+2 exists, we can say that it is logically possible to do that sum, and that, whether we actually do that sum or not, 2+2 always implies 4. I am saying, therefore when an object exists, there is also the existence of all the other objects that could be found by applying some automated interpretation process to it, whether it is run or not and regardless of what we mean by words like "cause, "because" and "imply" (which I feel is the best word for this).

Steve said: "I'd place it alongside things such as Zeno's Paradoxes."

But I still say that your position cannot easily be formalized, if at all - and mine can.

"I don't know how to answer that. Complex in which of several possible senses?"

I suppose in the sense that we would normally use to assess how well a description of a model satisfied Ockham's Razor - however you are right: everyone disagrees over what measure of complexity to use anyway, so it was probably not a good choice of question by me unless we want to spend a long time digressing into measuring complexity.

I want to say something about these other universes. It might seem that I am saying, because some extreme interpretation may be made of my chair to give Albert Einstein's brain that I am saying that there is a multiverse consisting of "sensible" reality, as well as all these crazy interpretations of low measure objects. Such single-step, extreme interpretations, however, would not be the interesting ones. A more interesting interpretation might be to make some interpretation of my chair, maybe even of some particle or group of particles in my chair at a single instant, to get a description of some laws of physics and a space-time, make more interpretations of that to get the basic particles in this "universe" and so on, so that an entire universe can be built up by applying interpretations. By the time you get to someone like Einstein living in that universe, his mind would be implied from the matter in his own brain by means of a much more conventional interpretation, just as ours are. Each individual step in such a staged interpretation would be much less extreme that what is needed to pull Einstein's mind, without a universe for him to live in, out of my chair. You can therefore make a series of interpretations that gradually find an entire universe in any object. I would say it is this that really makes this a multiverse model. You could then make interpretations of bits of that universe to find universes that it implies and so on, forever. There would be no reason to think that our universe is at the bottom of such a system, or that there is any bottom.

The idea of pulling Einstein out of my chair, in a single-step interpretation, therefore relates to a rather atypical, low measure thing, however I will briefly discuss this because it does have some, maybe, interesting strangeness to it. The Einstein found in such a way would have memories and experiences of working in the patent office, not knowing that all of his experiences were found in my chair. Even while he sits in his chair thinking about the theory of relativity, and observing the hands on his wall clock move, he is unaware that all of this experience is merely implied by my chair - possibly even by an interpretation of a single particle in my chair at a single instant. Does this mean that this Einstein's universe does not exist? Does it mean that he is a disembodied mind, adrift in the universe and sadly deluded? Well, not quite, because a description of the rest of the universe that this Einstein is supposed to inhabit could be extracted by some interpretation of his mind - and the fact that Einstein's mind will contain information that we would think of as corresponding to "proper" descriptions of the world he inhabits will actually reduce the contrivance needed to get some interpretations. In other words, the universe inhabited by such an Einstein would still exist - however his relationship with it would be the wrong way round. Of course, this works for you and I: the fact that our brains contain memories of things that have happened in this universe makes it less contrived to make interpretations of our brains to set up similar kinds of universes.

When we consider this, we should consider the issue of measure. Any universe implied by ours, or by some part of ours, will be given much less measure than our own universe. This does not mean that universes like ours have much less measure over all: they can be implied by other objects apart from our own universe. Our own universe would be implied by some interpretation of something else and other interpretations of the "something else" from which our universe is derived would give other universes, some very similar to ours and some vastly different. It would make little sense to discuss how many copies of our universe there are, or how common it is locally. All we can really discuss is measure in some more abstract, statistical sense. Some things would be easier to deal with. We would expect, for example, that almost all versions of Einstein's mind would come from universes where the existence of that mind would be a fairly sensible interpretation. The atom in the freakish state that you mentioned earlier would be much the same: most atoms really belong in universes. I can expect the Sun to rise tomorrow based on much the same kinds of reasoning. I can assume that my mental state is likely to be caused by a sensible interpretation of a universe which is itself the product of a sensible interpretation. Making universes where the Sun turns into cheese tomorrow morning would take a huge amount of interpretation, and making disembodies versions of my mind that experience that would be the same: it seems to me fairly obvious that the Sun will rise instead.

We have concentrated a lot on whether or not it is arbitrary to say that some things exist and not others, and that some interpretations are valid, or correspond to reality, and not others. This is only part of the case I have been making. We have not really discussed the other parts of this argument which I have briefly mentioned. These are:

1. The scenario of a conscious observer being encrypted in gradually stronger ways.
2. The issues of probability when identical minds, having identical experiences, are run on computers which are then split of combined - which I suggest destroys any possible notion of complete substrate independence in AI and leads to "statistical substrate dependence" and then, inexorably, to this sort of multiverse view. This argument was given, in part, in my articles on "Mind, Substrate, Measure and Value", on my website, where I described this sort of combination in detail and described a mechanical computer to show it - I posted them earlier in the discussion.

and a third issue which I have not mentioned yet:

3. The issue of probability when different copies of your mind are running on difference computers, with one copy being gradually encrypted as it runs - i.e. a combination of the ideas in (1) and (2).

Would you like to discuss these issues at some point? I suggest they are important to the case I am making and if you want to refute me they would need addressing, as the business we have been discussing is only part of it. I know you may not have time to read through all those articles, but I can summarize one of the issues here.

Hi Daniel

I am not really "looking" for any metric. I am quite happy with the idea that we do not need one, that we can simply base ontology on what is implied by objects, and that what is implied by an object is what can be produced from it by applying any interpretative algorithm to that object.

I am, however, demanding some kind of metric from anyone who wants to argue with me, so I will take this as a proposal that entropy satisfies the requirement. To apply a more extreme interpretation you would have to a lot more computation, and this would be associated with a lot more entropy, but I make the same objection that I have previously been making with other ideas for a metric: there is no cut-off point. How much computation or entropy is too much? How do you know that everyday objects you perceive around you, just by virtue of your brain being able to make the right interpretations do not have interpretations that violate whatever entropy condition you want, and therefore do not really exist?

I would also point out that basing this on entropy would seem to be basing it on the number of computational steps involved in the interpretation. This could lead to some strange results. For example, some object may be findable by a very small interpretative algorithm that needs to run through a huge loop of logical steps, running up a bit entropy bill, which could suggest (if we have some cut-off point) that it does not exist - despite the small finding algorithm. There is already a definition of complexity based on the number of computational steps needed to generate an object, known as "logical depth" and what you are proposing seems very much like a metric based on the logical depth of the interpretative algorithm needed to find an object. My own view is that the length of the algorithm is a better choice, as this statistically affects measure, and that this does not allow us to declare objects not to exist but merely means there are fewer of them around.

Hi Paul,

regarding
"I am, however, demanding some kind of metric from anyone who wants to argue with me, so I will take this as a proposal that entropy satisfies the requirement. To apply a more extreme interpretation you would have to a lot more computation, and this would be associated with a lot more entropy, but I make the same objection that I have previously been making with other ideas for a metric: there is no cut-off point."

I totally agree, one can not really say that some - let me say - incarnation of a phenomenon exists or not. However there is an important point with the idea of entropy: you really have to "sacrifice" something with low entropy in order to make a phenomenon appear in a certain way. As long as you cannot effort enough of something with lower entropy (e.g. electric potential), I would really say that this certain incarnation does not exist.

For instance if you have a bible and a computer but not enough electric power to decrypt it into becoming pornography, I would personally say this pornography in the bible really does not exist in this moment! It only comes to existence when adding enough electrical power. This of course does not make the pornography being less existent in the bible itself than the story of Genesis. I would actually say that both do not exist as long as one does not bring up the required increasement in (global) entropy. It is only that we usually bring up enough to bring the story of Genesis to existence, but not enough to bring the pornography into existence. And thus again I think that there still is a difference between both.

I really agree on "that this does not allow us to declare objects not to exist but merely means there are fewer of them around." however.

I was thinking about IKEA chairs too – I just mentioned rocks because a rock can be sat on and thus “becomes” a chair if we so choose to call it one, despite not looking much like furniture. I’m sure you’re right that there is some algorithm for recognizing chairs by appearance, although it’s by no means a trivial one, nor a conclusive one. We can recognize a chair by sight but largely because the concept has some meaning for us – we unconsciously ask ourselves whether it affords us the opportunity of sitting in it (as compared to eating it, having sex with it, etc.), and the answer decides whether it is a chair. But this need, this ability to recognize something’s affordances, could be imitated computationally even if the computer can’t itself sit down, so in that sense love seats, park benches and bar stools would continue to be members of a certain set of patterns after the human race was wiped out, just as many clouds would continue to resemble what used to be called faces. But I think what we really mean by “chairness” is something more humanly attributed. I don’t want to harp on about furniture though.

> Rather than considering my view idealistic, I consider this the most important feature of it: it demands that any position about what is real and what is not real can be formally described, at least in principle.

You realize I meant “Idealistic” with a capital “I”, as in the philosophical stance that there is no external reality without a mind to perceive it? You seemed to be requiring an observer, whether human or in terms of an operating algorithm, for things to be real. But you tell me this isn’t so. I’ll come back to that when I get to your comments on quantum measurement.

WHY must reality be formally describable? There have been many theories that had the advantage of being formally describable whilst simultaneously being wrong. Reality is what reality is (Idealism notwithstanding) and we may or may not be able to circumscribe it. Philosophy is a means of enquiry, not a legal requirement. Perhaps the truth is complex. Perhaps it is even arbitrary. Perhaps its lawfulness is beyond the descriptive capacity of logic or mathematics, since we know both to be incomplete. Certainly it is whatever it is without caring a jot whether we can formalize it.

So far it seems to me that you have a logically valid set of steps, albeit with an alarming number of “in principles” among them. But I can’t see that it gets us anywhere. I’m sure I’m just being dim but you seem to be saying that pretty much everything is real. That’s certainly a formally describable theory, but it’s a bit all-inclusive, isn’t it? If nothing fails the test then it isn’t much use as a test.

> You have to be able to say which are real and which are not, and I think you should be able to produce an algorithm which could do that on your behalf

But why? Certainly I think the term “real” isn’t worth anything unless it implies a converse. If everything counts as real then the word is useless. It’s like defining “red” so widely that a bluebell becomes a red flower with red leaves, contrasting not at all against a red sky. I think the term “reality” needs to have utility. Some things exist and others don’t – water supports sinusoidal ripples, vortices and a few other metastable configurations but no others – there must be reasons for that. Science is about trying to discern those reasons. But it doesn’t follow that we can necessarily form a single theory – a single algorithm, in your terminology – to separate what is real from what is not within all media at once. I can’t say what is or isn’t real, nor what will or will not be real in the future; I can only make a stab at it. The most I can say with certainty is that the universe is one way and not another – that some phenomena exist while others don’t. Any definition of reality that fails to admit that there’s a line to be drawn is a waste of a good word. Whether we can or cannot actually decide where to draw that line is another matter.

Ok, back to quanta:

>Why should something have more "right" to be regarded as real just because it is found by a simpler interpretation? I see no support at all for this philosophical rule.

I didn’t say it was a rule, but it’s certainly a useful rule of thumb. It’s not the fact that it’s a simpler interpretation per se, though, but the fact that a single, relatively simple interpretation can be made for each element of the series 1, 5, 7, but a massively more complex “algorithm” is suddenly needed to “find” element 3. Surely that makes element 3 look like an anomaly? If you did some experiments and the results formed a straight line graph, except for one point, which was many orders of magnitude out of alignment, would your first thought be that the function you’ve uncovered is highly nonlinear, or that it is linear and you made a mistake about that one point?

To put some flesh on the example, suppose these states were the quantum energy levels of an ion. Take an ionized gas (in a star, say) and take some simple “finding algorithm” for discerning facts about the population of ions (say you look at their absorption spectrum). You’ll find a characteristic large number of ions in a state I’ll call 1, another large group in a state I’ll call 5 and others in state 7. (I chose arbitrary whole numbers so that we could postulate intermediate states without being mislead by the fact that they’d have state numbers with decimal fractions). Now, my spectrometer will singularly fail to find ANY ions in states other than 1, 5 and 7 (or higher). It will never find any evidence of an ion in state 3. Moreover, by heating a gas I can demonstrate that ions in state 1 become ions in state 5 if they gain energy, and those in state 5 change to state 7. The states form a series, and therefore it’s fair to say that a hypothetical state 3, which has some characteristic feature such as a vibration mode that is intermediate in form between that found in states 1 and 5, is not an arbitrary state but part of that same series.

But we don’t find any ions in state 3. None whatsoever. The spectrometer – my finding algorithm – sees none of them. You say they exist and we just need a different finding algorithm. Yet you’re demanding a special case, because the existing algorithm works perfectly well for the points either side. And this new finding algorithm will be, I suggest, a great deal more complex and a great deal more contrived than a spectrometer. You haven’t discovered anything – you’ve invented it.

I suppose you could assert that states 1, 5 and 7 are actually the anomalies and there are an infinity of intermediate states, 3.000, 3.001, etc. which could be found by your “better” algorithm. But show me what that algorithm looks like. Mine wasn’t arbitrary or contrived. A spectrometer is not designed to find only certain “truths” about atomic state. It’s designed to look to see what is really out there, whether we like the answer or not. It is clearly logically related to the task in hand but is clearly not prejudging the answers. But I submit that any method you use to “prove the existence” of ions of type 3 will be very contrived and radically different from the algorithm that works for the points surrounding it.

I submit that the simplest, most elegant explanation - the one with the most utility and the least contrivance - is that ions of type 1 and 5 really exist, while ions of type 3 do not. Especially if I can show some logic for why some states should be stable while others aren’t.

> Another, and possibly better reason, is that these states have a strong relationship and interconnection with all the other stuff of which we are aware in the world. The 1, 5 and 7 states are part of our physics, which relates them, by various means, to everything from cows to fire engines. This may seem to give the 1, 5 and 7 states some special status

Well yes, that’s the point. They have the special status of actually existing! I concur with you that this is all relative – that one phenomenon only exists because other phenomena also do, and that (in places at least) this wraps around so that B is ultimately dependent for its existence on A, just as A depends on B. But what’s wrong with that? Perhaps the universe has sensitive dependence on initial conditions and could have turned out radically different. But it didn’t.

In my example above it has nothing to do with the limits of the human brain. Ions in state 1 stop photons of a certain frequency from passing close to them because they absorb them. Ions in state 5 absorb other photons, but nothing absorbs the photons that would have to be absorbed by a hypothetical ion of type 3. Therefore these don’t exist. I didn’t pick a spectroscope because I knew it would give me the “right” answer but because I knew it had a logical relationship to the question. I didn’t need any interpretation beyond trivial observation. The fact that the discovered state of affairs is only true because of the existence or non-existence of other phenomena (including the principles behind spectroscopy) is not an objection.

> What I am rejecting is the idea that those criteria [Steve’s requirements for emergence] dictate what exists and what does not exist and you have not said anything to persuade me that those criteria have that much ontological power. I still do not see from where such ontological power is supposed to come.

But what ontological power does your own theory have? It just admits everything into the reality club. As far as I can see, things either do or do not exist, whether we are able to know this or not. Reality is a given; it may or may not have a logic to it but it is what it is. All we have control over is our definition of it, and like all categorical descriptions it’s best if we choose a definition that has utility – one that divides candidates into members and non-members of the set. I was proposing some rules of thumb for identifying the process of emergence, not a hard-and-fast philosophically watertight definition of reality. Your definition seems watertight but looks like it fails to have any utility. It only has any use as a term if we admit that some things are more real than others, and I’m already happy to do that. It’s just that the vast majority of things you admit into the club seem to me to be so “unreal” as to be irrelevant for the purposes of understanding the universe.

Just because you can define an algorithm “in principle” for finding any pattern in any other, it doesn’t mean all algorithms are equal, or that there’s a smooth continuum between the everyday ones and the special cases you need . It seems to me that some “algorithms” are rational, uncontrived and meaningful, and then there’s a huge discontinuity before you get to ones that can find Albert Einstein in your chair or an ion of type 3 in a plasma. And in this discontuinity seems to lie a dishonesty – the algorithms for finding non-obvious patterns seem likely to be designed with one answer in mind.

Try it: describe a method of finding (or interpreting) something really simple that I say doesn’t exist, like an electron with a charge of -0.5.

> It would be almost certain that the observable universe - everything we know - is just an object that can be found by some interpretation of something else. This means that even our laws of physics, and our space-time would just be mathematical constructs that could be made from some other object by
some interpretation.

So it’s turtles all the way down, right? Didn’t you just complain that my interpretation of reality suffers from everything depending on everything else? Now you’re doing the same, aren’t you?

Just mathematical constructs? So is mathematics the foundation of reality? Very Pythagorean! :-)

> Maybe you think I am suggesting that observers or computation make things real? I am not suggesting at all that some observer, or anything else, needs to run some sort of machine or computational process to make an interpretation to make things real.

Yes, that’s what I thought you were requiring. I’m glad you’re not, but it still seems to me that you prize abstract human reasoning over evidence, somehow. I don’t have any problem defining “observer” in quantum theory – as far as I’m concerned it just means “interaction”. I’m relieved to hear you don’t have any truck with the “consciousness collapses the wavefunction” nonsense either.

> When I say that a thing exists because a particular interpretation can "find" it, that does not mean that I think something must actually make that interpretation

Ok, so everything is real purely by virtue of the fact that it COULD be “found”? You don’t actually need a measuring device. You don’t even need an algorithm. All you require is that some computational function could, in principle, construct one abstract “pattern” (whatever that means) out of another. That alone is proof that this other pattern exists?

I take it the nature of this computational function is irrelevant? Or does it have to be an algorithm? If every pattern implies every other pattern then surely it doesn’t matter how you derive it? You’re going to end up with the same answer anyway. Does this chair imply Albert Einstein? Yes. Does the chair that this implied Albert Einstein is sitting on also imply Albert Einstein? Yes. If the answer is always yes then the mechanism doesn’t matter, does it? Except perhaps to prove that the answer is never “no”? Are you able to do that? So the chair implies Einstein because I can merely visualize such a transformation as plausible? God exists because it’s possible to say he does? Does every finding algorithm imply the existence of every other finding algorithm?

I confess I just end up feeling like the universe is disappearing up its own backside with all this. Everything exists; everything implies everything else; everything depends on everything else except for mathematics, despite not needing any mathematicians. So there’s nothing that doesn’t exist, even things that we can’t conceive of. Nothing is only itself. Atoms imply universes which imply other atoms; everything that is not an atom implies an atom, so even non-atoms are atoms. Some algorithms are statistically more equal than others, it seems, but since all of them are only notional, “in principle” things anyway, there’s no way I can tell which is which. You seem to be portraying a multiverse in which nothing can be predicted, nothing can be said to not happen. And yet in my universe I can make predictions. Things that don’t happen don’t happen. The answers to my predictions are what they are whether I like it or not – I can’t just reinterpret them so that they give me the answer I wanted. Your multiverse might be logically consistent but it doesn’t seem to me to have any relevance to the universe in which I live! As my first wife would say, “what’s it got to do with the price of fish?”

Unless it gets us somewhere with regard to consciousness. Tell me about the incremental encryption of consciousness. I find myself wanting to retort that consciousness simply doesn’t work if you encrypt it, even if it exists in an encrypted universe.

Hi Steve

I will answer some of your points later, but for now, here is a scenario about gradual encryption of consciousness. I think other people may give different answers to this than either of us.

Fred is an uploaded human consciousness. His mind has been copied into a computer program from his brain, so that he can cheat death. The government has declared that all uploaded minds must be destroyed so Fred decides to hide.

Fred's first step is to stop interacting with the outside world. No more business meetings, charity auctions or online video games with other people: from now on, Fred will interact only with a simulated, virtual reality world (and we need not worry about whether the virtual reality world is real - this isn't what this is about). The police are searching computers for illegally uploaded minds, however, so Fred also tries to get the computer hidden in a safe place. Then he tries to get his code, and the code of the virtual reality world, hidden in an inconspicuous place on the computer.

This may not be enough. Fred wants to ensure that even if the police find the computer, and look at the right part of its memory, they do not notice him, so he applies some alteration to his code to make it slightly less obvious that what is running is a conscious mind in a VR world: you can see the program, but it is just hard to work out what it is doing. This is effectively encryption, but it is not major encryption: it just serves to mean you need a more careful look at the program to see what it is doing. Fred's first step is to use the most minimal encryption possible. Fred wants to be safer, so he improves his encryption again, and yet again - each time making himself slightly harder to find. The extent to which he becomes harder to find at each stage is the smallest extent possible: the least he can do to hide more. Fred continues this indefinitely. In the early stages it is obvious Fred is there, and for a number of stages near the start any decent programmer can see that the program is running Fred. After more stages only a better than average programmer can see that the system is running Fred. After a while, only the smartest programmers, inspecting the code very carefully for days, could see that the program is running Fred. After more stages, no unaided human mind could see that the program is running Fred. A computer would be required to assist in the search and apply some algorithm to the changing pattern of 1s and 0s in the computer memory and its registers to find Fred. There is no limit to this: the sophistication of the measures needed to find Fred increases with no end.

What is Fred's fate? Is Fred committing suicide? If Fred is really conscious at some point, does that consciousness cease to exist during this process? If so, when? Does it abruptly cease to exist at some single step in the process? Does it cease to exist at the first step or some later step (I would find that one REALLY untenable - and I would be surprised if you went for it? Does it cease to exist gradually? If it ceases to exist gradually what does it mean, to Fred, to say that he has partly ceased to exist? Is he still there but less conscious?

(One comment I will make here: If anyone thinks that Fred ceases to exist at the very first encryption stage, this is the most minimal attempt to hide the function of the program that is possible. This is a tiny degree of encryption. As an example, two competent human programmers A and B may attempt to write the program and A's code may be clearer than B's. The first step is no more than going from A to B - from code that would get a good mark for "good structured design" to code that would get a slightly less good mark - yet as this continues things just get harder to understand in small steps, without any end to the process. Even the very first version of Fred, before the encryption starts, could be viewed as encrypted from the point of view of someone who had just seen a similar process with a “better written” Fred at the start of it. You also have the issue of even making sense of the arrangement of matter in the machine and relating it to 1s and 0s. What I am saying here is that the concept of a completely unencrypted Fred is meaningless.)

Hi Steve

This is one of two response to your post. The other post, called “Progressive Encryption of Minds” is more important, I think, and was made earlier.

Steve Grand said: “You realize I meant “Idealistic” with a capital “I”, as in the philosophical stance that there is no external reality without a mind to perceive it?”

Actually, I did miss that. However, my view is not idealistic in that sense either.

Steve Grand said: “You seemed to be requiring an observer, whether human or in terms of an operating algorithm, for things to be real. But you tell me this isn’t so. I’ll come back to that when I get to your comments on quantum measurement.”

I do not require an observer: I want a formally describable position and needing something vague like an “observer” would just make things worse. I am not applying standards significantly different to what most people apply. I am just applying them more general. For example, both of us would probably agree that a group of pixels on a computer screen arranged according to the equation X^2+Y^2=R^2 are arranged in a circle. They are arranged in a circle because you can apply that equation to them and see that it fits: you can extract that pattern. This does not mean that some conscious observer is needed to compute all the values of that equation, or some computer, or in fact anything at all. We should also be suspicious of my use of the word “because” in the above sentence as well. It is not really that the pixels are arranged in a circle because you can apply that equation: the equation itself is not part of any causal process that makes circles appear. Rather, if an observer happens to come along and applies that equation and finds a circle he/she/it would validly say that the circle is there. It is not essential to discuss the issue of why the circle is there, when the equation can be used to find it there: we know that circles are associated with objects which can be described by the relevant equation. Any formal description of an object can be represented by an algorithm (unless we bring non-computability into this which, probably, neither of us need right now and could be dealt with if needed). All I am doing is taking a (much) more general view on what types of description of an object are valid, and admitting objects into existence when the formal description of how they relate to the things underpinning them is very complex. This does not mean I am right: I could be wrong to take such a general view. AN objection to what I have just said here is that maybe circles aren’t real: rather than get into a debate about Platonic reality, etc, I would instead point out that I could have just as easily used a more obviously physical construction, such as a sheep or cow: I could have said that a tree exists whenever the particular mapping that exists between a tree and the underlying reality exists, and that we do not have any problem just assuming that in everyday life without looking for the person who makes the mapping “real” by computing it or observing it. It is just the same for your views on emergent properties, which I do not disagree with: even your views would still require some kind of mapping from the underlying reality to the emergent object, and we presumably do not need someone or something to “execute” it.

Steve Grand said: “WHY must reality be formally describable? There have been many theories that had the advantage of being formally describable whilst simultaneously being wrong.”

I think the issue here should be more about why should descriptions of reality, produced by humans, or other thinking entities be formal – and I should have probably expressed it that way. If I produce a description of reality and it is formally expressed then, whether it is right or not, it is at least apparent what is being described. If I produce an informal description of reality then it must follow that more than one formal description of reality maps onto this informal description: otherwise it would be a formal description itself. Such a description is vague because it is not obvious what it is saying. I could define a language in which one word maps onto a million English words, but using that language to describe reality would be a bad idea because it would not be obvious what was being said in any description. That is my issue with informal views of reality vs. formal ones. I regard informal ones as vague so that they could describe many different situations. This is not just abstract philososphy: it has some practical implications. In the “progressive encryption of minds” thought experiment that I previously posted, without a formal description of the relatinship between a mind and the substrate, and what is valid for the substrate of a mind, it it is impossible to say what happens to someone’s mind, yet I would find it unreasonable to think that maybe it does not matter what happens to the mind, or that an accurate statement of what happens should somehow be beyond us, even in principle. There is also the issue of statistics and substrate, in the articles I mentioned, where you can run into problems while considering the combination of different computers running the same observer’s mind in a VR environment and have to determine the probability that you are in a given situation. I regard both of these as serious problems. I am not saying we have to be able to describe everythign formally, right now: I certainly cannot do that. I am saying that we should adopt an approach to ontology which at least allows that in principle. Further, the accusation that some idea is vague and beyond formal description is something I would use on many theistic views: without the requirement for formal description, or at least the glimmer of hope of one in principle, anyone can say anything.

Steve Grand said: “But I can’t see that it gets us anywhere. I’m sure I’m just being dim but you seem to be saying that pretty much everything is real. That’s certainly a formally describable theory, but it’s a bit all-inclusive, isn’t it? If nothing fails the test then it isn’t much use as a test.”

Yes, but with some qualifications. I am not saying that everything is real, but that everything that can be formally described is real – though I doubt you will find that any better. I am also not saying that everything is equally real. Some things are “more real” than others in the restricted sense that a greater proportion of patterns/mappings/algorithms generate them. It does not mean that individual occurrence of one of these objects is more or less real than any other though. Whether this really means “more real” is debatable, and largely semantics. It does mean that some objects would be represented in such a multiverse much more than others.

Steve Grand said: “Science is about trying to discern those reasons. But it doesn’t follow that we can necessarily form a single theory – a single algorithm, in your terminology – to separate what is real from what is not within all media at once.”

No, but I would say that the basic ontology used should at least allow it in principle.

Steve Grand said: “I can’t say what is or isn’t real, nor what will or will not be real in the future; I can only make a stab at it. The most I can say with certainty is that the universe is one way and not another – that some phenomena exist while others don’t. Any definition of reality that fails to admit that there’s a line to be drawn is a waste of a good word. Whether we can or cannot actually decide where to draw that line is another matter.”

Steve Grand said: “But we don’t find any ions in state 3. None whatsoever. The spectrometer – my finding algorithm – sees none of them. You say they exist and we just need a different finding algorithm. Yet you’re demanding a special case, because the existing algorithm works perfectly well for the points either side. And this new finding algorithm will be, I suggest, a great deal more complex and a great deal more contrived than a spectrometer. You haven’t discovered anything – you’ve invented it.”

Your spectrometer could indeed be viewed as a finding algorithm. It has some formal description and finds things that could be defined by a particular algorithm. I would say that it is still arbitrary though. A new finding algorithm may be more complex, but I do not accept that this makes it less valid. We could find something simpler than your spectrometer and use the same argument to say that your spectrometer is contrived. Of course, you can see your spectrometer, and what it is doing, and the things in the world with which it is engaging – but all of that is patterns found by your brain. This would bring us back to the basic issue of why you should think that all the things that your brain extracts from reality as “real” are “real” when all you have to go off is a pattern extraction. Even the spectrometer is a pattern extraction itself.

Steve Grand said: “I suppose you could assert that states 1, 5 and 7 are actually the anomalies and there are an infinity of intermediate states, 3.000, 3.001, etc. which could be found by your “better” algorithm. But show me what that algorithm looks like. Mine wasn’t arbitrary or contrived. A spectrometer is not designed to find only certain “truths” about atomic state. It’s designed to look to see what is really out there, whether we like the answer or not. It is clearly logically related to the task in hand but is clearly not prejudging the answers. But I submit that any method you use to “prove the existence” of ions of type 3 will be very contrived and radically different from the algorithm that works for the points surrounding it.”

The issue of “ions of type 3” is one that I don’t think would have much impact in reality. We would only call something an “ion” due to it interacting with reality in a certain way. If something was so removed from everything else, by resulting from some contrived interpretation, that it was literally just an ion in some state, with nothing else to interact with, it is debatable that that kind of semantics would even make sense. What we would probably have is some fairly boring, abstract pattern, ripped out of any meaningful. You would see it has no reality. I would say it is real. Neither of us are likely to be particularly interested by it. On the other hand, finding that pattern for an entire set of physics and building up from that would give a potentially more satisfying existence.

Steve Grand said: “I submit that the simplest, most elegant explanation - the one with the most utility and the least contrivance - is that ions of type 1 and 5 really exist, while ions of type 3 do not. Especially if I can show some logic for why some states should be stable while others aren’t.”

As I said, this is not really about trying to suggest that there is somehow an extra, meaingful state: I have no interest in trying to show that objects that have their meaning in context can have any interesting existence, or even make much sense, out of that context.

Steve Grand said: “In my example above it has nothing to do with the limits of the human brain. Ions in state 1 stop photons of a certain frequency from passing close to them because they absorb them. Ions in state 5 absorb other photons, but nothing absorbs the photons that would have to be absorbed by a hypothetical ion of type 3. Therefore these don’t exist. I didn’t pick a spectroscope because I knew it would give me the “right” answer but because I knew it had a logical relationship to the question. I didn’t need any interpretation beyond trivial observation. The fact that the discovered state of affairs is only true because of the existence or non-existence of other phenomena (including the principles behind spectroscopy) is not an objection.

It is not just about limits of the human brain, but about your own situation in reality. There may be some interpretation of some part of your world that produces some set of physical laws, a space-time continuum, observers, etc, but you would not observe it because the interpretation would be one beyond the capabilities of your brain to easily make. However, any of the observers in this system would easily be able to observe it, even if they were no more intelligent than you, because they would be in the right ontological situation. Their brains would be put together according to the same extreme interpretation as everything else around them.

This does not just have to apply to things resulting from our own world by extreme interpretation. If a view like this were correct, our world would probably just be an interpretation of “something else”. Other interpretations of that “something else” would produce other worlds, but these would be forever beyond our experience. We would just not be in the appropriate ontological situation to make observations of them,

Steve Grand said: “But what ontological power does your own theory have? It just admits everything into the reality club.”

I suggest it has greater consistency than a theory which arbitrarily admits things into the reality club. Further, it is subject to statistics, on account of this measure issue.

There are also some thought experiments involving observers in various strange situations. One type of situation is when an observer is being “run” on multiple computers which are combined. This is what I discuss in those “Minds, Substrate, Measure and Value” articles on my own website. I do not know if you have answer to the issue raised in those articles yet, but it is difficult to deal with minds and computers without some idea of measure, at least with regard to minds – this type of multiverse view is needed to make sense of that. Then there is the issue of the observer who is progressively more strongly encrypted, which I have posted here. I suggest my view deals with that too.

Steve Grand said: “As far as I can see, things either do or do not exist, whether we are able to know this or not.”

I am not really going to argue with that.

Steve Grand said: “It seems to me that some “algorithms” are rational, uncontrived and meaningful, and then there’s a huge discontinuity before you get to ones that can find Albert Einstein in your chair or an ion of type 3 in a plasma. And in this discontuinity seems to lie a dishonesty – the algorithms for finding non-obvious patterns seem likely to be designed with one answer in mind.”

Again, things like making an extreme interpretation of my table to find Albert Einstein in his chair is not what this is really about: though I would be dishonest if I did not admit that such interpretations are possible, and, in my view, would have to be accepted as corresponding to real objects. I simply take an ontological view which happens to produce things like that, rather than start to cut and slice where I think I need to to be left with what serves our purposes. When it comes down to the particulars of generating models, and making predictions I am likely to function more or less in the same way as you.

I can give one example where there would be no such discontinuity: the example of “Fred” who gets progressively encrypted, which was in my last post. Each time Fred makes the encryption process slightly stronger, the finding algorithm needed to find him gets slightly longer.There can be a whole continuum of progressively more encrypted Freds. You don’t just have the “sensible” Freds, and then a desert in the set of finding algorithms, before more Fred’s start to appear at 2^1,000,000,000 bits.

Steve Grand said: “So it’s turtles all the way down, right? Didn’t you just complain that my interpretation of reality suffers from everything depending on everything else? Now you’re doing the same, aren’t you?”

Depends. I admit I am still thinking through some aspects of this.

One argument would say that if a view like this were correct, it would be very unlikely that anything we could encounter would be the “bottom level” and that it would not make sense to say there is one, from our point of view.

Another view would say that the entire system is founded on interpretation of nothing. The first interpretations would be based on nothing as an object. They would be applied to nothing and would generate the first layer, which would then be interpreted to provide the next layer and so on: there would be an infinite number of layers. This would mean that there would be a bottom level, in principle: a set of interpretations which had no input at all, but it then raises the issue of where we would expect to be in such a situation. If we expected to be an infinite number of layers from the bottom level then, as far as we were concerned there would not be one.

When I said your view involves everything depending on everything else, and my view is the same, I disagree. I admit everything, which means there are no arbitrary decisions about what is admitted. Your view involves decisions about what is admitted which are based on other decisions and so on.

e.g. A is real because it is a consequence of B being real. A is real because B is real.

It is this arbitrary, mutual dependence on what is admitted into reality with which I am taking issue.

Steve Grand said: “Just mathematical constructs? So is mathematics the foundation of reality? Very Pythagorean! :-)”

I would say mathematics is the foundation of any well developed ontology that describes reality. Whether it is the foundation of reality itself is semantics.

Steve Grand said: “Ok, so everything is real purely by virtue of the fact that it COULD be “found”? You don’t actually need a measuring device. You don’t even need an algorithm. All you require is that some computational function could, in principle, construct one abstract “pattern” (whatever that means) out of another.

Yes.

Steve Grand said: “That alone is proof that this other pattern exists?”

Yes – although the argument itself is not just based on saying that selecting particular patterns are arbitrary. There are also thought experiments on things like encrypting observers, or combining computers which are simulating observers and I think it best deals with these kinds of situations.

Steve Grand said: “I take it the nature of this computational function is irrelevant? Or does it have to be an algorithm? If every pattern implies every other pattern then surely it doesn’t matter how you derive it? You’re going to end up with the same answer anyway. Does this chair imply Albert Einstein? Yes. Does the chair that this implied Albert Einstein is sitting on also imply Albert Einstein? Yes. If the answer is always yes then the mechanism doesn’t matter, does it? Except perhaps to prove that the answer is never “no”? Are you able to do that? So the chair implies Einstein because I can merely visualize such a transformation as plausible?”

Steve Grand said: “God exists because it’s possible to say he does?”

No. You might define some being that manifests itself in the world in much the same way that God does in, for example, the bible: burning bushes, prophets, etc. You might even imagine a being that creates a universe. Providing this can be formally described, it would be out there in the multiverse somewhere. Someone, somewhere, would be living in a universe with prophets, messiahs, etc and which had been created by some being. This would not necessarily make it common. It might seem strange that a cosmology like this says that things like this would exist, but it not much worse than a universe with an infinite, or even just very large, amount of space. While science does not tend to suggest these days that there is an infinite size to the universe, it would hardly be absurd in any deep sense if it were the case, and if it were we would probably all accept that there are really low probability, silly things out there somewhere, but so far away, and so infrequently occurring that they are just statistical anomalies of reality rather than anything of any real interest. Lots of things would be found in a cosmology like the one I suggest here, but that does not make them important: effectively this is just like a big cosmos where the ontological relationship between things is different. There is an important point about any apparent gods that are found in this way: they would not go any way at all to meeting the theistic idea of “God”. They would be beings that were emergent from something else in reality, due to being “found” by some algorithm. God is not supposed to result from anything or be based on anything. The inhabitants of whatever world they have made may think they know everything, have made everything and are all powerful, but in reality they would bevery provincial beings, not deserving the title of God with regard to anyone else in reality. They would also be formally describable – not something that theists tend to advocate. For example, Alvin Plantinga, a Christian apologist insists (absurdly in my view) that God is “basic”.

With regard to the frequency with which these “beings who look like Gods but aren’t really” occur, that would be a matter for more discussion and I think two people could agree with this kind of cosmology and have different views on it.

Steve Grand said: “Does every finding algorithm imply the existence of every other finding algorithm?”

Yes, because every object could have an appropriate finding algorithm applied to it to find every other object, which could in turn have every other finding algorithm applied to it and so in. In practice, when it comes to any serious attempt in making predictions most of these would not concern us.

Steve Grand said: “I confess I just end up feeling like the universe is disappearing up its own backside with all this. Everything exists; everything implies everything else; everything depends on everything else except for mathematics, despite not needing any mathematicians. So there’s nothing that doesn’t exist, even things that we can’t conceive of. Nothing is only itself.

Emergence does some of that anyway: all this does is much more of it. That objection ignores the issue of measure. Objects which tend to be findable by short algorithms will be found a lot more often than objects which need contrived finding algorithms. Things would be quite statistically predictable. In fact, a view like this can be used to justify something like Occam’s razor. When determining the probabilities of things you hardly need to consider all of reality: you would really just need to consider the “local” ontology as it is that which has the main influence on the statistics of what is happening around you.

Steve Grand said: “Atoms imply universes which imply other atoms; everything that is not an atom implies an atom, so even non-atoms are atoms.”

Not the language I would use. You and I would both say that atoms, in the right arrangements, imply trees, but we would not say that atoms are trees. Similarly, I would say that non-atoms imply atoms, rather than saying they are atoms.

Steve Grand said: “Some algorithms are statistically more equal than others, it seems, but since all of them are only notional, ‘in principle’ things anyway, there’s no way I can tell which is which. You seem to be portraying a multiverse in which nothing can be predicted, nothing can be said to not happen. And yet in my universe I can make predictions. Things that don’t happen don’t happen. The answers to my predictions are what they are whether I like it or not – I can’t just reinterpret them so that they give me the answer I wanted.”

Steve Grand said: “Your multiverse might be logically consistent but it doesn’t seem to me to have any relevance to the universe in which I live! As my first wife would say, ‘what’s it got to do with the price of fish?’”

I think that the “what’s it got to do with the price of fish?” argument is a bit weak, because that could be said about any philosophy in this kind of area, or at this kind of level. You could say the same about views about whether or not computers have minds, whether a machine that can exhibit apparent intention has real intention, what a mind is, what emergent properties are, etc. For example, the entire issue of whether or not Searle is right could be dismissed with an appeal to fish pricing relevance. I doubt that much, if anything, of this kind of issue was discussed, for example, during the Manhattan Project – and I doubt absence of it was an obstacle.

Steve Grand: “Unless it gets us somewhere with regard to consciousness. Tell me about the incremental encryption of consciousness.

Since you asked this, I have now provided that in my previous reply to this – the hypothetical of Fred being encrypted.

Steve Grand said: “I find myself wanting to retort that consciousness simply doesn’t work if you encrypt it, even if it exists in an encrypted universe.”

That won’t work. There is no point at which something suddenly becomes “encrypted”. The way a mind is run on some physical system can be made gradually more complicated, equivalent to encrypting it progressively more strongly, without any line being passed at which it suddenly becomes “encrypted”. The example of progressive encryption of Fred, in the post before this one, should show that.

I wonder if the left margin is going to keep moving to the right?