Several regulars have requested that I put together a short OP and I’ve agreed to do so out of deference to them. Let me be clear from the outset that this is not my preferred course of action. I would rather discuss in a more interactive way so that I can learn from criticism and modify my thoughts as I go along. OPs are a little too final for my tastes.
I want to emphasize that everything I say here is tentative and is subject to modification or withdraw as feedback is received,
It’s important to understand that I speak for no one but myself it is likely that my understanding of particular terms and concepts will differ from others with interest in ID. I also want to apologize for the general poor quality of this piece I am terrible at detail and I did not put the effort in I should have due mainly to laziness and lack of desire.
With that out of the way:
Background
For the purpose of this discussion I would like to expand upon the work of Phill Mcguire found here and stipulate that cognition can be seen as lossless data compression in which information is integrated in a non-algorithmic process. The output of this process is a unified coherent whole abstract concept that from here forward I will refer to as a specification/target. Mcguire’s work thus far deals with unified consciousness as a whole but I believe his incites are equally valid when dealing with integrated information as associated with individual concepts.
I am sure that there are those who will object to the understanding of cognition that I’m using for various reasons but in the interest of brevity I’m treating it an an axiomatic starting point here. If you are unwilling to accept this proviso for the sake of argument perhaps we can discuss it later in another place instead of bogging down this particular discussion.
From a practical perspective cognition works something like this: in my mind I losslessly integrate information that comprise the defining boundary attributes of a particular target; for instance,”house” has such information as “has four walls”, “waterproof roof”, “home for family”, “warm place to sleep”, as well as various other data integrated into the simple unified “target” of a house that exists in my mind. The process by which I do this can not be described algorithmically. from the outside it is a black box but it yields a specified target output: the concept of “house”.
Once I have internalize what a house is I can proceed to categorize objects I come across into two groups: those that are houses and those that are not. You might notice the similarity of this notion to the Platonic forms in that the target House is not a physical structure existing somewhere but an abstraction.
Argument
With that in mind, it seems reasonable to me to posit that the process of design would simply be the inverse of cognition.
When we design something we begin with a pre-existing specific target in mind and through various means we attempt to decompress it’s information into an approximation of that target. For instance I might start with the target of house and through various means proceed to approximate the specification I have in my mind into a physical object. I might hire a contractor nail and cut boards etc . The fruit of my labor is not a completed house until it matches the original target sufficiently to satisfy me. However, no matter how much effort I put into the approximation, it will never completely match the picture of an ideal house that I see in my mind. This is I believe because of the non-algorithmic nature of the process by which targets originate. Models can never match their specification exactly.
Another good example of the designing process would be the act of composing a message.
When I began to write this OP I had an idea of the target concept I wanted to share with the reader and I have proceeded to go about decompressing that information in a way that I hoped that could be understood. If I am successful after some contemplation a target will be present in your mind that is similar to the one that exists in mine. If the communication was perfect the two targets would be identical.
The bottom line is that each designed object is the result of a process that has at its heart an input that is the result of the non-algorithmic process of cognition (the target). The tee shirt equation would look like this
CSI=NCF
Complex Specified Information is the result of a noncomputable function. If the core of the design process (CSI) is non-computable then the process in its entirety can not be completely described algorithmically,
This insight immediately suggests a way to objectively determine if an object is the result of design. Simply put if an algorithmic process can fully explain an object then it is not designed. I think this is a very intuitive conclusion, I would argue that humans are hardwired to tentatively infer design for processes that we can’t fully explain in a step by step manner. The better we can explain an object algorithmically the weaker our design inference becomes. If we can completely explain it in this way then design is ruled out.
At some point I hope to describe some ways that we can be more objective in our determinations of whether an object/event can be fully explained algorithmically but as there is a lot of ground covered here so I will put it off for a bit. There are also several questions that will need to be addressed before this approach can be justifiably adopted generally such as how comprehensive an explanation must be to rule out design or conversely when we can be confident that no algorithmic explanation is forthcoming.
If possible I would like to explore these in the future perhaps in the comments section. It will depend on the tenor of feed back I receive.
peace
A quick note. I have previously come across this approach to cognition within the AI community. I’m extremely skeptical of it. But I’ll take the time read the cited paper before I comment further.
And thank you for posting. You have already cleared up what you meant by “lossless”, and given me a better picture of what you are talking about.
fifth,
That’s exactly why we don’t stop with the OP. We keep going in the comment thread! I guarantee that you’ll get a lot of interaction and feedback.
This is a question rather than a critique. Your cognition seems to be essentialist.
I have designed a number of things. Call them micro-inventions.
Those that have worked have not resembled my original vision. In fact, some don’t even do what I originally intended them to do.
fifth,
I think you need to tighten up your thesis a bit.
You would regard a Shakespearean sonnet as an example of CSI, correct?
Specify any sonnet and I can easily create a program that will produce that sonnet.
I can also easily create a program that will produce all of Shakespeare’s sonnets (along with a ton of gibberish) simply by exhaustively producing all character strings up to a given length.
I think the two processes being mutually exclusive is a good falsification criterion?
keiths says,
You would regard a Shakespearean sonnet as an example of CSI, correct?
I say,
Have you been looking at old UD threads?
keiths says,
Specify any sonnet and I can easily create a program that will produce that sonnet.
I say,
Would the program completely explain the sonnet?
peace
petrushka says,
Those that have worked have not resembled my original vision. In fact, some don’t even do what I originally intended them to do.
I say,
define “worked”?
First of all, thank you for your pdf. I have long thought that AI is a much more difficult problem than most people think.
What do I mean by worked? I can discribe a recent micro-invention.
I have a project to digitize about 50 years of family photos from negatives. My estimate is about 4000 images on negatives. I have several negative scanners, but they are really slow when attempting to get high resolution. Recently I purchased a vintage slide copier attachment having a universal camera mount. With an adapter, it attached to a DSLR.
The problem is it requires a slide mount, and I have negatives. The target is something that allows passing a strip of negatives through a device designed for individual slides.
To keep a boring story as short as possible, I will only say that after about 50 attempts I wound up with a slide mount with a couple of tiny modifications.
In retrospect the object is obvious. In construction it is a point mutation.
petrushka says.
The target is something that allows passing a strip of negatives through a device designed for individual slides.
I say,
Did the final product match your target closely enough to satisfy you?
Peace
Yes. The scanners took about 10 minutes per image. With the DSLR I can digitize about 200 an hour.
Please understand at this point I am not arguing against you. Just exploring the concept of target and design.
Criticism:
After the start of the “Background” section all lines are centered, rather than left-justified, as is done properly in the previous paragraphs.
Will be more readable if someone can fix that.
petrushka says.
Yes.
I say,
Then we have nothing to disagree about so far.
It’s possible that as I go about the process of designing a house I will find that the color of the siding or the size of the windows will change the floor plan might even be radically altered. This sort of thing happens as our plans meet reality
But the final result if I’m successful will match my original target (house) well enough to satisfy me.
Peace
The pdf has math that is over my head, but I think I get the point about integration of information.
Here’s the metaphor I use personally when thinking about memory. I think of my life as a fixed resolution image, zooming out like one of those videos that start with a person on the street and back away until you see the whole earth.
My memory starts at or near birth and zooms out to include 70 years of experience. But the number of neurons is roughly constant. As you age, the view gets wider, but you can’t just do a CSI enhance. Detail is gradually lost in grain or pixelation. Compression is not lossless.
Joe says,
Will be more readable if someone can fix that.
I say
The powers that be have my permission to modify as they see fit. It all looked much better before I pasted it from open office
peace
petrushka says
Compression is not lossless.
I say
I would hope you could accept my axiom for the sake of argument, I’d hate to get bogged down so early in the discussion.
I would be willing to discuss the conclusions of the paper at a later date. The math is tough but it’s not necessary to understand the gist.
peace
fifth,
No, but I remember them.
keiths:
fifth:
Not sure what you mean by “completely explain”, but remember, your criterion was whether the sonnet could be produced algorithmically. It definitely can.
keiths said.
but remember, your criterion was whether the sonnet could be produced algorithmically. It definitely can.
I say,
Not sure if I said that at some point but it’s definitely not my criteria now. I agree that any finite string can be produced algorithmically. But would disagree that a sonnet can be explained by the algroythym that you use to produce it.
To explain a sonnet you need both an approximating algorithm and the target you specify. You might want to check out the other thread on active information for details.
peace
I accept that memory is integrated. I have thought since I first became aware of science fiction AI, about 50 years ago, that brains do not store information the way desktop computers do, and that attempts to download and upload consciousness will fail. I’m not claiming to be correct on this, just saying I have believed it for a long time.
That is not at all the same thing as saying brains are not physical objects with components similar to computer components. I have no objection to physical minds.
What do you suppose would be the target of an e.e. cummings poem?
petrushka.
I have no objection to physical minds.
I say,
I can’t say I do either. I haven’t thought about it much. The point is not that the mind is immaterial just that cognition is non algorithmic
peace
For some definitions of algorithmic. Keiths and I went round and round on this.
fifth,
My point is that every finite string can be produced algorithmically, so if you take “cannot be produced algorithmically” as your criterion for design, then you’ve ruled out every novel, every poem, and every computer program.
I’m pretty sure that’s not your intention, so that criterion won’t work.
But if that’s not your criterion, then what is?
keiths thinks that if an algorithm can produce an output (what algorithm doesn’t produce an output), that saying the algorithm produced the output is an explanation for the output that was produced by the algorithm.
keiths:
Perhaps you should go back and read the OP.
I would tend to say cognition is an evolutionary algorithm.
fifthmonarchyman,
How are cognition and consciousness related?
Would you please remark on the panpsychism of integrated information theory?
The response I expect here is that something has to generate the program. There’s a universal dovetailer that interleaves execution of all programs, i.e., a program that runs all programs. [Edit: It’s certainly not designed to hit a target.]
I’m not sure I’m a Power That Be but it did let me edit your post, so that is now done.
Thanks, fifth! This gives us something to discuss, which is great! It does not need to be your final thoughts on the subject 🙂
Tom English asks,
How are cognition and consciousness related?
I say,
I would say consciousness is more general awareness and cognition is awareness in a more restricted particular sense.
You say,
Would you please remark on the panpsychism of integrated information theory?
I say,
If I was not a Christian I would probably hold to panpsychism of some sort I find it to be appealing at many levels.
As it is I think that in integrated information theory there tends to be a confounding of what consciousness does (integrate information) and what it is.
That is not unique to that field. I’ve seen some here who seem to have the same trouble deciding if the mind is what the brain does or a real thing of itself.
peace
Just keeping this thread on the leader board.
In my discussion of targets I never got around to stating what I think a biological target is. So here it is.
A biological target is a variation that results in equal or better reproductive success.
A biological target is not a new structure or function. Those things happen by serendipity, but cannot be targets, because nothing in reproduction has foresight. Success in being equal to or better is hindsight.
http://www.sciencedaily.com/releases/2015/05/150512075107.htm
The analog computer thing is something I’ve been saying for several decades. It could be wrong, but I feel a bit less alone in the universe now. It’s relevant to the “integration” and “non-algorithmic” arguments.
This went round the other day: http://phys.org/news/2015-05-pattern-recognition-magnonic-holographic-memory.html perhaps of interest.
Dunno. Not convinced that analog is the key.
fithmonarchyman,
Thank you for clearly indicating that you’re engaged in apologetics. In particular, you told us that your assumption of determinism and rejection of panpsychism were theological. In integrated information theory, even inanimate objects are conscious to a low degree. Tononi readily accepts that this is panpsychism. You openly explain — thank you — that you reject it because you’re Christian.
I am not claiming to discern hidden intent in what you’ve written. You’ve identified positions that you hold on religious faith, and begun drawing what is convenient (and rejecting what is inconvenient) from IIT to construct an argument that makes a “conclusion” out of what you’ve already decided must be true. That is apologetics.
I appreciate your openness about your beliefs, and about your intentions. Do please continue to post at TSZ. I’ll point out, however, that your comments on the post created for you, after you expressed interest in active information, were almost all off-topic. I won’t appreciate it if you’re constantly injecting your ideas into threads where they don’t belong. TSZ, unlike UD, allows you to post OPs. So it’s not too much to ask that you behave differently here than there.
I don’t think the brain is any kind of computer – analog or digital.
If by no kind of computer you mean has not been copied or emulated by any inorganic computer, I’d agree.
It depends on what you understand by “a computer”. The brain hasn’t got a neat von Neumann architecture and may be analogue rather than digital, but — like a computer — it is a general-purpose information processor which exchanges vast amounts of information with the outside world. Our artificial computers are barely more than half a century old. Not very brain-like yet, but ask again in one hundred years.
This is mean, but I think there’s a bit of cargo cult in AI. Copying the outward form rather than the principles.
I guess I was not clear enough.
I don’t think the brain is any kind of information processing system.
I hope that’s clearer.
Tom English says,
Thank you for clearly indicating that you’re engaged in apologetics.
I say.
I often engage in apologetics. The term after all simply means to offer a defense. I’m not however doing so right now. I’m simply trying to firm up some ideas I have.
You say,
In particular, you told us that your assumption of determinism and rejection of panpsychism were theological.
I say,
We all have various reason for holding the beliefs that we do. Are my reasons of lessor value than yours?
you say,
You’ve identified positions that you hold on religious faith, and begun drawing what is convenient (and rejecting what is inconvenient) from IIT to construct an argument that makes a “conclusion” out of what you’ve already decided must be true. That is apologetics.
I say,
I don’t think that word “apologetics” means what you think it means.
And your characterization of my motives and method are incorrect as well. I have no idea if my speculations are correct I haven’t even settled on them completely, I thought I’d made that clear in the OP.
you say,
I’ll point out, however, that your comments on the post created for you, after you expressed interest in active information, were almost all off-topic.
I say,
I’m sorry but I happen to believe that a discussion of the limits of evolution with regards to targets is on topic when discussing active information.
Oh well germane comments are in the eye of the beholder I guess.
you say
I won’t appreciate it if you’re constantly injecting your ideas into threads where they don’t belong.
I say
Just let me know if you don’t approve of something I say I will be happy to drop it and move on. I’m a little slow some times but I promise I don’t want to offend you in any way. I only came here on a whim in the first place
peace
This might be of interest as well when thinking about the differences between computer learning and the model of cognition I am talking about.
http://www.evolvingai.org/fooling
peace
Is there a working implementstion of your model?
petrushka says.
Is there a working implementstion of your model?
I say.
I have a tool that I use to help determine if a particular “object/event” can be explained algorithmically. That’s all
I’m not sure if it meets the threshold of scientific model but it’s way cool IMHO. It would take a lot of time to explain it all. It exists in a crude spreadsheet format.
I’m trying to learn some rudimentary programing so that I can share it at some point. And perhaps work on some hypothesis testing in a more controlled setting .
peace
fifth,
Still awaiting your reply to my earlier question:
It appears that the imáges doing the fooling were evolved.
keiths asks.
But if that’s not your criterion, then what is?
I say.
From the OP
This insight immediately suggests a way to objectively determine if an object is the result of design. Simply put if an algorithmic process can fully explain an object then it is not designed.
Just because any finite string can be “produced” by an algorithm does not mean it can be “explained” by said algroythym. Like you alluded to earlier if I have a target I can produce an algroythym to approximate it.
To “explain” a designed object you need the target as well as the algroythym. I hope we can flesh this out as we move forward.
peace
petrushka says,
It appears that the imáges doing the fooling were evolved.
I say
Yes evolution is very good at that sort of thing. I hope you don’t think I’m anti evolution.
peace
fifthmonarchyman,
You do not understand IIT. You might get commenters on your posts — plural — to work along with you to achieve a better understanding. But your style of interaction does not encourage that. I recommend that you go with a post, “What Is Integrated Information Theory?” You might outline the field, acknowledge difficulties that you have with it, and invite people to work with you toward a common understanding. Then follow through:
The notion that you achieve this better by munging threads is garbage.
By the way, your responses would be much more readable if you used
<blockquote> quoted text </blockquote>
P.S.–Just back from a wonderful dinner with the “fundie” family.
fifth,
Set the target equal to a specific Shakespearean sonnet. Then the algorithm “Print target” successfully produces a string matching the target.