🚀 go-pugleaf

In article <35d9e013.9366758@news.demon.co.uk>, bob@lintilla.demon.co.uk
(Bob Cousins) wrote:

> In comp.ai, Rugrat wrote:
> 
> >Tony Johnson wrote in message <35B397A7.BF6F010D@juno.com>...
> >>Mentifex wrote:
> >>> Somewhere before or beyond the Millennium lies a crack in time, a
> >>> fork in the ontology of conscious and intelligent beings on Earth.
> >>
> >>-describe to you're self in your own words, what the difference is you
> >>observe between the colors blue and red...
> >
> >In your own words? So who here has a word they can call their own?
> >
> >Blue and red, firings of relevant cell arrays in LGN correspond closely with
> >people's reportage of how they perceive the whole spectrum of colors. ie. We
> >see these colours because our brains are made to see these colours. Human
> >beings, irrespective of culture, we make similiar reports on the colour
> >spectrum. The difference can now, even if still approximately, be discerned
> >by the way our brain differentiates the light spectrum.
> >Don't chase ghostly qualia, the neural firing is the percept.
> 
> Blue is a concept we learn to associate with light of a particular wavelength
> (unless you are colour deficient), but that doesn't mean that blue *is*
light of
> a particular wavelength. Colour is a shared subjective experience, but
> subjective all the same. A blind person may know about blue, but if you gave
> them sight, would not be able to identify it without an example.
> 
If color is a subjective experience, then everything is subjective and the
word objective loses all meaning.  In fact, color detection is one of the
basic information our sensory systems provide; this is the basis for all
empirical knowledge. You should trust it more than mechanical
instrumentation, which, ultimately, is calibrated to someone's sensory
experience (however that person(s) decided to calibrate it). 

Plus you are pushing the learning part of this too much, especially with
respect to color. Color is fairly hardwired in the nervous system and
there is not that much genetic variability in it either (and I am learning
psychologist). There are probably better examples, but not color..


> The sensory input just provides us with a fairly unique set of signals to
> associate with other things. The implication is that an AI will need senses
> similar to ours, in order to have an intelligence we can relate to. But even
> then, it will not derive by itself the same concepts that we have. It could be
> that intelligence as we define it is purely a subjective view of how we
develop
> to react to sensory input. An AI based on logic, may perform useful tasks, but
> may never seem intelligent.
> 
> Anyway the difference between blue and red is that blue is a cool peaceful
> colour, whereas red is a warm vibrant one.