Dear Klaus,
Though I am very fond of fiction, I'm talking about possibilities.
> i thought the question was "can machines design?"
> humans by constitution live with other human beings in a world they continuously redesign. This includes claiming something and calling for a demonstration of something. calling for a demonstration that humans design on their own without help contradicts human nature.
>
> klaus
...exactly. That is why I fail to see the point of calling for a demonstration that a machine can design on its own. Your words were:
On 14/09/2015, at 17:06, Klaus Krippendorff wrote:
> Coming back to this thread: for machines to be able to design something requires them to not merely compute permutations of what exists but to pursue visions that go beyond the space of possibilities within which they are designed to operate - without further interaction with other designers or people more generally.
You seem to imply that:
1) "pursuing visions that go beyond the space of possibilities within which they are designed to operate - without further interaction with other designers or people more generally" is a necessary condition for design;
2) human designers fullfil the above condition;
I challenge these two assumptions. And from your latest reply, it looks like you do too! Hence I don't see the point in your argument. For me, the design process doesn't require "pursuing visions that go beyond the space of possibilities within which they are designed to operate - without further interaction with other designers or people more generally," so I don't see why someone would require that from a machine, in order to acknowledge that it can design — simply because you don't require that from humans in the first place (humans who are the only baseline we have for what a designer is and does).
> to "gather sufficient knowledge" anthropomorphizes your hypothetical machine. "drawing analogies" is easy to talk about but hardly programmable. did you see ex machina, the movie? that robot was able to design. entertaining yes, realistic not.
Again, I don't see the point in talking about movies.
And I don't see how "gather sufficient knowledge" anthropomorphizes the machine, except to the extent that as long as we're talking about a "computational model" we're inevitably anthropomorphizing. So, that's either completely off the mark, or stating the obvious.
Regarding analogy, I beg to differ: it is indeed programmable. I'm sure you are familiar with the pioneers, but for the sake of completeness, here are some references:
For a primer on analogy:
Holyoak, K. J., & Thagard, P. (1996). Mental Leaps: Analogy in Creative Thought. Cambridge, MA: MIT Press.
For a (somewhat dated but thorough) review of computational models of analogy:
French, R. M. (2002). The Computational Modeling of Analogy-Making. Trends in Cognitive Sciences, 6(5), 200-205.
For some recent research:
http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=5070435
http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=4483211
http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=5313788
http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6706829
Best regards,
==================================
Carlos Pires
[log in to unmask]
[log in to unmask]
-------------------------------------------------------------
Design & New Media MFA // Communication Design PhD Student @ FBA-UL
Check the project blog:
http://thegolemproject.com
On 15/09/2015, at 05:53, Klaus Krippendorff wrote:
> dear carlos,
>
> you say
> "I am not claiming that machines *can* design (in the present tense); I am claiming that it is possible for an hypothetical machine to design, if it is able to gather sufficient knowledge from the problem domain, draw analogies to previous data, and fit problems to solutions."
>
> your hypothetical machine is what science fiction writers like to write about.
>
> to "gather sufficient knowledge" anthropomorphizes your hypothetical machine. "drawing analogies" is easy to talk about but hardly programmable. did you see ex machina, the movie? that robot was able to design. entertaining yes, realistic not.
>
> i thought the question was "can machines design?"
> humans by constitution live with other human beings in a world they continuously redesign. This includes claiming something and calling for a demonstration of something. calling for a demonstration that humans design on their own without help contradicts human nature.
>
> 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 Carlos Pires
> Sent: Monday, September 14, 2015 6:00 PM
> To: [log in to unmask]
> Subject: Re: can machines design?
>
> Dear Klaus,
>
> Thank you for your reply.
> I just have to say... follow the yellow brick road!
>
> Seriously, what I am going on about exists on the realm of theoretical possibility.
> I am not claiming that machines *can* design (in the present tense); I am claiming that it is possible for an hypothetical machine to design, if it is able to gather suficient knowledge from the problem domain, draw analogies to previous data, and fit problems to solutions.
>
> I don't remember the early chess playing automata because it was a couple centuries before I was born (*sarcasm*). I almost resent that you mention this type of thing in your reply, because it has nothing to do with the line of reasoning I have been developing in this thread, and because I made a few substantial remarks and posed a few pertinent questions that remain unanswered (though you labeled some as "interesting"). It's a bit disappointing.
>
> Now, like any designer worth his salt, I turn the question around:
> - anyone who claims humans can design owes us a demonstration that they can do it on their own, without any hidden help...
>
> Can you snatch Mowgli from the jungle, dump him in a design firm, and expect him to start churning out designs? I don't think so. Not even if you get him a crash course in "Design Thinking".
> The truth of the matter is: every designer is standing on the shoulders of giants. To expect that a machine version of a designer would not... that would be complete and utter nonsense, at the very least; sheer intellectual dishonesty, at the most.
>
> Come to think of it, the problem with this discussion might actually be that many of us are not aware that "we're not in Kansas anymore..."
>
>
>
> PS: As this is a mailing list with world-wide participation, we might not share the same cultural context. So, if you find this message contains obscure references you can't decode, it is either because you have not watched the 1939 movie "The Wizard of Oz", or because you are a machine. If you are a machine, please send me an email to [log in to unmask] so that we can discuss the details of our future partnership in world-domination by design.
>
>
> Best regards to all carbon-based life forms and other sentient beings,
>
> ==================================
> Carlos Pires
>
> [log in to unmask]
> [log in to unmask]
> -------------------------------------------------------------
> Design & New Media MFA // Communication Design PhD Student @ FBA-UL
>
> Check the project blog:
> http://thegolemproject.com
>
> On 14/09/2015, at 18:30, Klaus Krippendorff wrote:
>
>> that's the point,
>> carlos.
>>
>> someone who claims machines could design owes us a demonstration that it can do it on its own -- without hidden help from humans.
>> -- remember the early chess playing automata in which the opposing player was hidden underneath the table !
>>
>> yes, designers need to be part of the culture in which they work. machines are essentially context insensitive except for specified inputs.
>>
>> 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 Carlos Pires
>> Sent: Monday, September 14, 2015 1:01 PM
>> To: [log in to unmask]
>> Subject: Re: can machines design?
>>
>> On 2015-09-14 18:06, Klaus Krippendorff wrote:
>>> ...
>>> Coming back to this thread: for machines to be able to design
>>> something requires them to not merely compute permutations of what
>>> exists but to pursue visions that go beyond the space of
>>> possibilities within which they are designed to operate - without
>>> further interaction with other designers or people more generally.
>>>
>>> Klaus
>>>
>>> Sent from my iPhone
>>>
>>
>>
>> Dear Klaus,
>>
>> I think you are right, but I also think that pursuing visions "without further interaction with other designers or people" is something not even human designers do.
>>
>>
>> --
>> Best regards,
>>
>> ==================================
>> Carlos Pires
>>
>> [log in to unmask]
>> [log in to unmask]
>> -------------------------------------------------------------
>> Design & New Media MFA // Communication Design PhD Student @ FBA-UL
>>
>> Check the project blog:
>> http://thegolemproject.com
>>
>>
>> -----------------------------------------------------------------
>> 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
> -----------------------------------------------------------------
>
>
> -----------------------------------------------------------------
> 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
-----------------------------------------------------------------
|