Hi, Terry,
To speak of “machine-created” designs gets into ontological questions I am not ready to address. As I wrote earlier, I am going to think these issues through to explain why I maintain the position that machines cannot design in the sense that humans can design. I’ll argue that machines cannot “create” for much the same reason, though for a different range of issues.
It isn’t possible to deal with all questions at once. Some of the questions here aredistinct from the issue of using insights to develop design principles and even distinct from how to develop principles from design research findings. The issue of whether machines can design came up only because you changed the text of the Cooler Solutions advertisement. Your changes to the text emphasized formal methods. This created equivalence between the human being that Cooler Solutions seeks and the notion that machines can design.
Cooler Solutions seeks a senior-level applied social scientist “versed in all phases of design research: from proposal strategy, research design, conducting research, development of insights and translating them into design principles.” Achieving insights requires meeting and interacting with clients and problem owners to understand their problems. Machines still have problems understanding ordinary language, let alone understanding the troubled, ambiguous, and frequently emotional language involved when professionals diagnose problems. Designers must select and solve the “real” or “key” problem among many ambiguous possibilities. That’s the issue I was trying to bring forward in my earlier note, and my claim is that there is no mathematically rigorous formal system for doing this.
As Herbert Simon said, design is partly heuristic. Design will always be partly heuristic because it is lodged in judgment and preference. Simon’s proposal for a design science is that some aspects of design and some aspects of the research that supports design can be rendered scientific and partially formal. Until machines can handle both aspects of design, they can’t design. But of course, machines don’t do science or mathematics either – machines support the human beings who do science and perform programmed calculations for mathematicians.
When I finish thinking my way through the notion of flat ontology, I will return to answer your post. This will take me a couple of months. I will defer the debate onwhether machines can design until I am ready to say something sensible and well founded.
Yours,
Ken
Professor Ken Friedman, PhD, DSc (hc), FDRS | University Distinguished Professor | Swinburne University of Technology | Melbourne, Australia | [log in to unmask]<mailto:[log in to unmask]> | Phone +61 3 9214 6102 | http://www.swinburne.edu.au/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
-----------------------------------------------------------------
|