On 2/23/2018 8:46 AM, Bruno Marchal wrote:
I have developed a fuzzy version of prolog, actually based on Philippe
Smets (the main creator of IRIDIA) “transfer belief model”, to help
in some search in giant protein databanks (when I was working for
Monsanto (!)). With a friend we have taken some years to compile it in
ADA. He told me that a subroutine we created has actually been
implemented in all chips (a sort of hash-table data retrieving
procedure! I am not sure if this can have value in AI, or perhaps I do
think a fuzzy prolog can have value, but I am not sure if this is
exploited today. At that time, I heard about a Japanese firm which
built a fuzzy chip for logic programming for a “clever” cleaning
machine. I believe more in neural nets, especially if it is cycling,
and with ¨many¨hidden layers. You can be sure that when such machine
will think, we will not understand how they do it,
And neither will the machine.
and such machine will be beyond theoretical analysis. With computers,
either we massacre the semantics, and they will be efficacious, but
often wrong and a bit stupid in the human way, or we will built docile
slaves. It is about the same with the education of kids. We will
forever remains in-between security and liberty ...
Yes, having raised four kids and now advising (to no effect) how to
raise grandkids and great-grandkids I sometimes reflect on how the No
Free Lunch theorem may apply to education.
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
To post to this group, send email to firstname.lastname@example.org.
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.