terry,
you are asking me "Are you saying that human abilities in thinking, feeling and intuition are not biologically limited?"
no, i am not claiming that. i go beyond this trivial claim by saying everything that we know of and have experience with has limitations. i cannot jump over my own shadow, my computer cannot store more data than to its capacity, we humans cannot live long without food.
also, i need to remind you that biology is a science. it develops biological theories. these theories may explain but do not limit being human. and they are subject of frequent revisions. so human beings may be bodily limited but not biologically limited.
historically, experiencing human limitations have stimulated the development of language (to collaborate), culture (to reproduce successful practices) and technology (to amplify our abilities or exerting forces, compute alternatives, search the world).
my problem is with your premise of machines being so much better (at designing) than human beings. obviously, who would deny that my search engine can go through a couple of million documents to search for a word i am interested in. i would probably take years to do that. who would deny that a machine gun does more damage than i can do with bare hands. the point is that machines are designed, designers invented their "rules", we have evolved. machines can be designed to adapt to changing conditions but this is not the same as inventing or changing one's own rules. you are dreaming of artificial intelligence but haven't ever addressed the artificial stupidity of machines.
i know of parallel processors that learn but only if someone programs the criteria for what constitutes a better fit. i know "self-organizing" networks but they have no clue about the rules they follow.
this is why i suggested to be epistemologically clear and distinguish between an observers theories of how something works and a human being inventing or adopting a set of rules and enacting them knowingly -- at least for some time -- like when playing bridge, or being a design researcher at one point and a parent at another.
klaus
-----Original Message-----
From: Terence Love [mailto:[log in to unmask]]
Sent: Saturday, September 19, 2015 3:42 AM
To: Klaus Krippendorff; 'PhD-Design - This list is for discussion of PhD studies and related research in Design'
Subject: RE: can machines design?
Hi Klaus,
Are you saying that human abilities in thinking, feeling and intuition are not biologically limited?
PS. I agree with you that it is important to ask who specifies the rules (and why they chose those rules).
PPS: On self-generation of rules, Kohonen networks and other networks with learning ability have this competence as do most self-educating robot systems. At a larger scale, the Watson system manages it - even including inappropriate language.
Best wishes,
Terry
-----Original Message-----
From: [log in to unmask] [mailto:[log in to unmask]] On Behalf Of Klaus Krippendorff
Sent: Saturday, 19 September 2015 12:02 PM
To: PhD-Design - This list is for discussion of PhD studies and related research in Design
Subject: RE: can machines design?
a comment about "rule-based".
i think you have to always aske who stated the rules?
in the case of machines, a programmer spells out how a computer behaves (which does not imply that all of their behavior can be anticipated). you know whether a system is rule-based by asking its designer -- until it breaks down.
in the case of the human brain, those who claim that the brain follows particular rules (e.g. of logic, etc.) are mistaken. cognitive scientists tend to theorize the rules a brain is said to follow base on observational evidence. historically, such theories have been revised and continue to be subject of revision which is sufficient evidence that the brain is not based on the rules that observers have theorized.
the same is true for dna. biological explanations of how dna works have evolved in the discourse of biologists. dna does not know these rules, are not affected by them, and cannot follow them either. they would have to study biology.
incidentally this applies also to terry's claims of human limitations. they are derived from an irrational privileging of mathematical and machine explanations. these limitations reside in terry's use of language, not in humans.
human behavior can be rule-based, when it concerns behavior that language can govern, e.g., playing soccer, complying with the laws of a country, making commitments and living up to them. in social reality, this applies only to a small part of social life.
indeed, if you could program designers, then they could become rule governed. i doubt that we have the rules of successful design practices and teach them to merely receptive students.
" What happens to any rule-based entity, if the first rule is: "define your own rules" ?" -- first this entity needs to have the capacity to understand and use language constructively -- what you are asking.
second this entity needs to be able to create a space of possible rules and chose among them -- minimally choose to change from following one rule to another.
third that entity needs to be able to follow their own rules (subordinate future choices to the rule adopted, potentially suspending the ability of this entity to define its own rules). humans do that all the time. after following the rules as a soccer player, following traffic rules while driving a car, preparing a meal by following a recipe, .... for human beings, this is not to extraordinary.
what would be extraordinary if terry could design a machine that demonstrates all of these abilities.
klaus
-----Original Message-----
From: PhD-Design - This list is for discussion of PhD studies and related research in Design [mailto:[log in to unmask]] On Behalf Of Joran Booth
Sent: Friday, September 18, 2015 7:17 PM
To: [log in to unmask]
Subject: Re: can machines design?
Thank you for your questions, Carlos. Here are my answers.
1) What evidence is there that humans are not rule-based themselves?
Frankly, I cannot think of any evidence right off. However, I consider certain conditions to indicate non-rule-based behavior.
First, emotion seems to fit this category since it is non-rational by definition. Intuition also seems closely related.
On the other hand, one could argue that DNA is ultimately a long series of rules with a certain deterministic probability. Can see this point, or similar ones, and I would be unable to argue against someone who sees it this way.
2) What happens to any rule-based entity, if the first rule is:
"define your own rules" ?
This is an interesting question indeed. "Define your own rules" seems to imply that there is a presupposed ability to do so. If defining rules is a rule based activity, the logic becomes circular, and therefore programmable. However, if the ability to define rules is not logical, but instead arbitrary, I would suppose this would mean that we cannot. Do you have another take on it?
However, I wonder if it is possible for something to rise above its own rules anyway. It could be that an agent cannot construct levels of abstraction higher than the one it operates on. (Something like abstraction entropy, maybe?)
On 9/9/2015 10:33 AM, Carlos Pires wrote:
> On 2015-09-09 16:01, Joran Booth wrote:
>> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256
>>
>> This is a really interesting thread.
>>
>> Here are my two cents:
>>
>> 1) Machines are inherently rule-based. Even heuristics are rules.
>> To my knowledge, there is no machine that has been created yet hat
>> does not have a rule-structure underlying it. 2) Design is design
>> precisely because it is NOT rule based (Jonassen 2000, Toward a
>> Design Theory of Problem Solving). 3) Novice designers rely heavily
>> on rules, whereas experts do not (Dorst 2004, The Problem of Design
>> Problems), so even if machines could design, they would be inferior
>> to human designers.
>>
>> Based on these points, I would conclude that machines cannot design.
>> Can they make beautiful works from very complicated algorithms, but I
>> would argue that this is not design. It is math (which is often
>> quite beautiful).
>>
>> - -Joran Booth Ph.D. Candidate Purdue University
>>
>
> Hi Joran,
>
> Interesting. Let's consider for the sake of the current debate that
> machines are rule-based. I have two questions, then:
>
> 1) What evidence is there that humans are not rule-based themselves?
> 2) What happens to any rule-based entity, if the first rule is:
> "define your own rules" ?
>
--
Joran Booth
Ph.D. Candidate
Purdue University
West Lafayette, IN
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]> Discussion of PhD studies and related research in Design Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]> Discussion of PhD studies and related research in Design Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]>
Discussion of PhD studies and related research in Design
Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
|
