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
fifthmonarchyman,
The output might be rendered as pixels on a screen, or printed to paper, or read outloud by text-to-speech conversion, but the output _is_ the sonnet.
By your own definition, the algorithm explains the target.
Clearly you have something else in mind, but until you provide an operational definition, no one else has any way to know what that is.
fifthmonarchyman,
I read the paper, but I don’t see how it addresses design detection. At most it shows that some assumptions about capital markets may be too simple. I’m quite confident that neural networks or SVMs could learn to distinguish between curves like those in the paper.
Why, exactly, do you think this paper supports your claim to be able to objectively detect design?
Patrick says,
I don’t think you understand what I mean by nonlossy compression.
Our brains integrate all the information involved into one unified lonlossy compression in this case “how to ride a bike”.
We can’t just change one small part of that information and expect the compression to still work.
That is the point of the paper and the video demonstrates it quite nicely
peace
Patrick says,
I say,
It points to a way that we can distinguish between strings that is inaccessible to computers.
The paper was looking at real verses randomized data but I discovered the same method works with real strings verses algorithmically produced ones that are close.
peace
Patrick says,
I say,
If an algorithmic software system could learn to nonlossly integrate all the information in the string like we do I would have to reevaluate what I believe consciousness is. I think it would count as a strait up passing of the Turing test.
It would cause me to completely reevaluate my entire approach. It would mean abandoning any hope design detection by this method.
peace
I’m beginning to see where we are miscommunicating.
As best I can tell, you take the brain to have memorized instructions for riding a bike, and to have compressed and stored those instructions into memory.
My view is that the brain has paid very little attention to said instructions. Rather, the brain has developed ways of measuring how well I am in balance when riding, and has grown feedback loops to generate corrective motions to keep me in balance. As I see it, learning is far more like tuning a piano than it is like storing data into a computer memory.
Here’s a thought experiment for you. Teach a 6-year old child to ride a bike.
Then, once he reaches age 7, he never touches a bike again until he is age 20. What happens?
My expectation is that if he tries riding at age 20, he will be very wobbly and have trouble keeping in balance until he relearns. And the reason for my expectation is that, as he has grown, the mass distribution within his body has changed greatly so that the old feedback loops are not going to be very effective.
Neil Rickert says,
I say,
I would probably agree with that expectation.
It does not affect my argument at all. Again I’m not saying humans never forget I’m just talking about how cognition works
peace
Neil Rickert.
I say,
Exactly, that it the essence of what I’m trying to get across when I say cognition is non-computable.
I really think we are probably allies here as far as the first part of the OP goes except for a few idiosyncrasies.
What I need you to do is take the next step (if you haven’t yet) and understand the connection between cognition and design. That is where the fun will begin.
peace
Patrick says,
Patrick says,
I say,
again from the OP
quote:
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.
End quote:
The target is the concept that exists in the author’s mind it involves much more than words on page or pixels on a screen.
You could translate the sonnet into Chinese or update it into modern Arabic. You could re-imagine it into a rap song set in twenty-first century Los Angeles. It does not matter the target remains the same.
peace
I think that I do understand the connection.
Here’s the connection, as I understand it: It is impossible to design a cognitive system. Even a god could not design one. You can only get a cognitive system by evolution. A truly omniscient god would know that, so would create evolution as the way to produce cognition.
Neil Rickert says,
I say.
interesting take. Not exactly the one I was going for.
I would probably say it’s impossible to design a cognitive system that functions algorithmically.
you say,
I say,
How can a algorithmic process produce something that is beyond the capabilities of algorithms?
You say,
I say,
again interesting but you need to explain how you get there from here
Peace
It’s a way scientific endeavour and theism could rub along together, Scientists get on with discovering God’s universe and theists marvel at his omniscience. Now if theists could just rein in their tendency to tell other people how to live their inner lives…
Oh!
You appear to think that evolution is algorithmic. You have probably been reading too much Dawkins.
I see evolution as adaptive rather than algorithmic.
It seems that they cannot. Too many of them are authoritarian control freaks.
If only they could be more adaptive.
Yes, if only …
If god could rein in its believers so that they quit trying to ruin our mutual world, I might even be motivated to join in the worship.
Sadly, the god they’ve got as far as we can tell has either no interest or no power to emend their hearts and dissuade them from being such nasty busybodies.
Not really a god worth much, after all.
Neil Rickert says
I say,
step one (random mutation)
step one A( plus drift, plus HGT, plus etc etc etc)
step two (natural selection)
Step three (repeat steps one through two indefinitely)
looks algorithmic to me what am I missing?
peace
Allen Fox says,
I have no problem theologically with Neil Rickert’s idea but I’m just not too concerned with segregating my thoughts into different spheres of influence. It just feels unnatural to me
I would rather pursue truth and let the chips fall where they may.
I do find it interesting that it seem to be folks on your side of the fence that are continuously bringing up the theological here.
A while back when I gave an honest theological answer to a direct question Tom English accused me of merely practicing apologetics here .
I wonder if he would reach the same conclusion when the theology comes from his side.
peace
Mutation is not algorithmic, and natural selection probably isn’t either.
Neil Rickert says,
I say
Individual steps are not “algorithmic” in any given process.
It’s the entire process that is algorithmic
Any time you have a step by step process you have an algroythym by definition.
I would hope we could agree that evolution is a step by step process
peace
You may have just reduced “algorithmic” to meaninglessness. Anything can be described as a step by step process. So, by your definition, everything is algorithmic.
Neil Rickert says
I say,
Not at all. there are lots of things (such as cognition) that are not algorithmic.
Some things happen all at once with no intermediate steps at all
Some things happen with no process at all they are effectively random
Some things are a like a black box in which we only know the inputs and the outputs. Knowing what if anything happens in between is forever beyond our grasp.
I’m sure there are other processes that are non-algorithmic but it’s late and I can’t think of them right now. But you get the idea
peace
Yes. And mutation and natural selection are good examples of that.
You ignored why mutation might be non-algorithmic, by just taking it to be a single step. So take anything to be a single step, and you have a step-by-step process (i.e. an algorithm).
That’s what you appear to be doing.
Normally, one use “algorithm” with mathematics or information processing. It’s a sequence of steps where every step is precisely defined and determined. Neither mutation nor natural selection is precisely defined and determined.
Hey Nick Rickert
I did not expect this discussion to be an philosophic exercise. I do hope we can avoid that. At this point it does not interest me too much. It’s usually my side that want’s to go down that road.
Neil Rickert says,
I say,
That may be so, I’m agnostic.
Evolution as a whole on the other hand is an algorithmic process by definition. In fact it meets the definition exactly.
from here
http://en.wikipedia.org/wiki/Algorithm
quote:
an algorithm is a self-contained step-by-step set of operations to be performed.
end quote:
That is exactly what evolution is.
you say,
I say,
wow
1) I wonder if your fellow critics would agree that mutation and natural selection are not precisely defined
2) That an algorithm’s individual steps must be precisely defined is not a part of the definition
3) Ever hear of Evolutionary Algorithms?
moving on
I want to understand you. Is this just going to be a claim about the existence of true randomness?
Are you just saying that the results of evolutionary processes are not predictable by us so therefore evolution is not an algorithm?
Would you agree that if an observer was omniscient and the universe is deterministic the results of evolution would be completely predictable and therefore in that case it is algorithmic?
Just trying to understand you
peace
I disagree.
From the second paragraph of your linked Wikipedia page:
I could not express evolution in a formal language, and I doubt that you could either. And evolution is not calculating any function, as far as I can tell.
No. It has more to do with the differences between physical reality and abstract mathematics.
Neil Rickert says,
I say,
Are you saying that EAs can not model actual evolution even in theory?
If that is the case how is evolution not equivalent to “whatever happens”?
peace
Evolutionary algorithms are approximations of the natural process; a model. One difference is that many events can happen simultaneously in nature.
ETA: All models are wrong, but some are useful. — George E. P. Box
That’s my opinion. Yes.
They can demonstrate some of the capabilities of evolution. But you cannot adequately model evolution without first modeling the actual environment.
Well, I can simply define “mutation” as a change in a gene. I think Neil means something precise by “precisely”. Similarly, with natural selection I can simply define the process as “change over time in allele frequency in a breeding population of living organisms”.
Exactly!
DEM can make no claims about the efficacy of evolution by producing a model of evolution that doesn’t model evolution and then suggest because their model of evolution cannot work, evolution is impossible.
ETA: well, they can (and do) but it’s not effective.
I hesitate at this point because I don’t want to sound like the stereotypical fundi with a Tu quoque argument and that is definitely not my intention.
But I’ll ask again in case you all missed it …………..
If that is the case how is evolution not equivalent to “whatever happens in the universe”?
Please be specific in your answers. I think this might be important moving foward.
If we can’t define what evolution is with any precision how can we test it? If we can’t test it how is it science?
Peace
Would there be any objection at this point if we were to stipulate that
If evolution turns out to be in any any sense an algorithmic (ie a step by step process) then it can not produce a cognitive function or what an observer would consider to be a designed object?
I just want to take a snapshot of where we are in this discussion.
peace
It’s hard to answer, because the question seems so wrong.
I suppose that “whatever happens” could be seen as the basis for the weak anthropic principle. But you don’t get very far with that. Evolution depends very much on biological reproduction. That biological reproduction is something of a driving force.
I sometimes say that I am not a Darwinist. And people seem to not understand that. Folk at TSZ have wondered why I say that, when it seems to them that I agree with a lot of what is neo-Darwinism.
As I see it, Darwinism is roughly the claim that biological systems are passively adaptive due to natural selection and chance mutation. However, my view is that biological systems are aggressively adaptive. Mutations are not just chance events — the production of mutations is part of the system and the mutation rate itself can evolve.
That’s certainly how I visualize the process, with the niche environment being the engine of selection.
Are we merely disagreeing about outcomes? The possibility of mutations, where and how often they occur can all be non-random but the outcome of any mutation is not predictable and the effect of any mutation can only be established in the context of the particular niche environment. Beneficial mutations are not just good, they are good for something – the organism in its particular niche.
“In any sense” is too weak. That makes it too easy to construe non-algorithmic systems as algorithmic.
From my perspective, as a mathematician and computer scientist, a computer operating system is not algorithmic though it uses many algorithms. It is not algorithmic because there is no target that it is trying to reach. It is supposed to just keep on keeping on (with apologies to Sherwin Williams). An algorithm is supposed to have a target and to halt when it reaches that target.
An algorithm proceeds according to strict rules of logic. And that’s why we build computers out of logic gates. But biological systems are composed of homeostats rather that logic gates. They don’t strictly follow rules. If a rule does not seem to be working, they replace that with a better rule.
In terms of decision making: an algorithm proceeds by using truth as the basis for all of its decisions. A biological system is fundamentally pragmatic. It makes its decisions based on what works rather than on truth. It does not depend on knowledge of ultimate truth, which is just as well because there probably isn’t such a thing.
So evolution is far more creative than any mere algorithm could ever be.
Hard to say.
I see biological populations as actively attempting to expand their niche.
That’s a teleological step to far for me. 🙂 It smacks of group selection when really, all individuals are just living.
I don’t see any implications of group selection there.
Teleology? Maybe. But teleology is just homeostasis dressed up in fancy clothes.
Your phrase I see biological populations as actively attempting to expand their niche suggests a kind of plan. Some group ethic to conquer their environment. What I see [is]* individual, and unequal, struggle. Maybe I read more than is intended.
As a thermostat’s behaviour is regulated by the environment, gene frequency is regulated by the environment. All individuals can do is survive and reproduce if they can. The changing environment settings regulate the change in allele frequency.
* oops
I’m not suggesting any group ethic. I was merely giving my overall description of the behavior. There does not need to be planning.
I don’t agree with that description. The thermostat provides a feedback loop. One might even say that it is a regulatory feedback loop. And that’s the source of the homeostasis and the apparent teleology.
Depends on what the thermostat is controlling. The output of the thermostat depends on the immediate ambient temperature around the probe. Which is what I mean when I talk about environmental design. Sure organisms in an environment can adopt strategies to control their environment. This generally leads to adaptations that reinforce those strategies. Rather a form of positive feedback. Take naked mole rats as an example. Did the rats design their tunnels or did the niche design the rats to reinforce the behaviours that led to social behaviour, tunnelling and all the parallel morphological adaptations.
ETA just noticed you did not say “negative feedback” so I may be at cross purposes here.
fifthmonarchyman,
That’s an incredibly idiosyncratic use of the term “non-lossy”, to the point of confusing communication.
Just because the way memory works in human brains is significantly different than how we store data digitally does not mean that human memory is lossless. As several people have pointed out in this thread, the opposite is demonstrably the case.
“Integrated information” conveys what you mean, I believe. “Non-lossy” is non-descriptive and a distraction. I hope you will agree to stop using it unless and until you can support the idea that human memory is non-lossy by more standard definitions.
fifthmonarchyman,
It does no such thing. It hints that one particular simplifying assumption made about capital markets may be inaccurate.
fifthmonarchyman,
Let’s be very clear here, because you’ve rewritten my question to bake in your own assumptions.
First, human memory is not non-lossy, as I noted again a couple of comments ago.
Second, my question is: If it could be demonstrated that a software program could learn the task described in the paper, how would that impact your argument? I said nothing about the software being able to “integrate all the information in the string like we do.” I assure you that it would not pass the Turing test.
Here is what I am specifically suggesting,
1. Create up to 1000 pairs of time-series data as described in the paper.
2. Create either a recurrent neural network that accepts the time series data directly or a support vector machine that takes in statistics directly generated from the data.
3. Train the network with 800 or so pairs of time-series data and validate it with the remainder.
4. Create a few hundred new pairs of time-series data for testing.
5. Demonstrate that performance on the test set meets or exceeds the human performance described in the paper.
I’d have to play with approaches in step 2 to find the quickest way to model the problem, solely because I don’t want to spend months on the project.
If my prediction in step 5 turns out to be true, would it result in you “abandoning any hope [of] design detection by this method”?
fifthmonarchyman,
That’s not how you defined “explain”, though. If you are claiming that an artifact under discussion is “explained” only when we take the causal chain back to a human mind, then your whole process is unnecessary. Just like Dembski’s CSI, it’s something that you ascribe only after confirming (human) design.
It also makes your process useless for detecting design in biology.
If this is not what you mean, please provide a new operational definition for “explain’.
From the actual paper
quote:
We suggest that such novel interfaces can harness human capabilities to process
and extract information from financial data in ways that computers cannot.
end quote:
geez
peace
Patrick says,
I say
I am using the standard definition. The authors of the paper also use the standard definition and go into detail to explain exactly what they mean. That you insist on reading some other definition when you see the term is your problem not mine.
I suggest you use the definition that the authors use when discussing the topic of the paper.
peace
Patrick says,
I say,
That is not my claim at all.
My claim is that a artifact is “explained” when it points you to the target. My original definition works just fine to convey this understanding. I make no mention of a mind human or otherwise.
The target is nothing but a nonlossy data compression. The information in the target is decompressed in the artifact.
Because Algorithms compress data lossily they are inadequate to explain artifacts.
Ive been through this and I would be happy to elaborate at any point if you would only ask
However I’m beginning to get the feeling that you are not looking for clarification at all but are instead only looking to score debate points.
It makes for a unfruitful discussion at this point of the process.
peace
fifthmonarchyman,
Regardless of whether or not evolution is algorithmic by whatever definition you and Neil might agree on, the empirical observations are that humans are one result of it. If we have “cognitive function” (another term for which you need to provide an operational definition) then evolution is demonstrably capable of producing it.
So yes, I have a strong objection to making stipulations that are not in accordance with the available evidence.
Patrick says,
I say,
once again if an algorithmically based process can nonlossily integrate all the information in a string in direct contradiction of the mathematical proof offered in the original paper I linked in the OP then yes I would abandon any hope of design detection by this method.
I won’t hold my breath
Now it’s possible that a a recurrent neural network might become proficient at distinguishing real financial data from randomized copies of that data but in that case I would predict that it will be easy to fool the network in the same way we can fool the visual systems in the third paper I linked.
Thus demonstrating that the network is not actually integrating the information the way we do.
long story short if your software can distinguish between strings and at the same time not be subject to being easily fooled by EAs I would abandon the whole enterprise.
How is that for a falsification criteria?
peace
Patrick says,
I say,
well there you go
“If it exists it will turn out be the result of evolution. Because we know that everything that exists in biology is demonstrably the result of evolution”
quite a tight circle to that logic you got going there.
peace