RE: UDA revisited

From: Stathis Papaioannou <>
Date: Sat, 25 Nov 2006 20:21:46 +1100

Colin Hales writes:
[All good, rousing stuff, I wish I had more time to reply!]
> I spent tens of thousands of hours designing, building, benchtesting and
> commissioning zombies. On the benchtop I have pretended to be their
> environment and they had no 'awareness' they weren't in their real
> environment. It's what makes bench testing possible. The universe of the
> zombies was the universe of my programming. The zombies could not tell if
> they were in the factory or on the benchtop. That's why I can empathise so
> well with zombie life. I have been literally swatted by zombies
> (robot/cranes and other machines) like I wasn't there....scares the hell
> out of you! Some even had 'vision systems' but were still blind.
> so....yes the zombie can 'behave'. What I am claiming is they cannot do
> _science_ i.e. they cannot behave scientifically. This is a very specific
> claim, not a general claim.
You're being unfair to the poor zombie robots. How could they possibly tell if
they were in the factory or on the benchtop when the benchtop (presumably)
exactly replicates the sensory feeds they would receive in the factory? Neither
humans nor robots, zombie or otherwise, should be expected to have ESP.
> >>- it's all the
> >> same - action potential pulse trains traveling from sensors to brain.
> >
> > No, it's not all the same. Its coded in a very complex way. It's like
> saying the information in you computer is "all the same -- its all ones
> and zeros"
> yes you got it - all coded....I am talking about action potential pulse
> trains. They are all the same general class. Burst mode/Continuous mode,
> all the same basic voltage waveform, overshoot, refratory period...LTP,
> LTD, afterhyperpolarisation.... all the same class for sight, sound,
> taste, imagination, touch, thirst, orgasm etc etc... coded messages
> travelling all the way from the periphery and into the brain. They are all
> the same...and..
> None of it says anything about WHY the input did what it did. The
> causality outside the zombie is MISSING from these signals. They have no
> intrinsic sensation to them either. The only useful information is the
> body knows implicitly where they came from..which still is not enough
> because:
> Try swapping the touch nerves for 2 fingers. You 'touch' with one and feel
> the touch happen on the other. The touch sensation is created as
> phenomenal consciousness in the brain using the measurement, not the
> signal measurement itself.
> Now think about the touch..the same sensation of touch could have been
> generated by a feather or a cloth or another finger or a passing car. That
> context is what phenomenal consciousness provides.
But it is impossible to differentiate between different sources of a sensation unless
the different sources generate a different sensation. If you close your eyes and the
touch of a feather and a cloth feel the same, you can't tell which it was. If you open
your eyes, you can tell a difference because the combined sensation (touch + vision)
is different in the two cases. A machine that has touch receptors alone might not be
able to distinguish between them, but a machine that has touch + vision receptors
would be able to.
> >
> >> The zombie cannot possibly distinguish the novelty from the sensory
> data
> >> and has no awareness of the external world or even its own boundary.
> >
> > Huh? It's perfectly possible to build a robot
> > that produces a special signal when it encounters input it has
> > not encountered before.
> Yes but how is it to do anything to contextualise the input other than
> correlate it with other signals? (none of which, in themselves, generate
> any phenomenal consciousness, they trigger it downstream in the
> cranium/cortex).
That's all we ever do: correlate one type of signal with another. The correlations
get called various things such as "red", "circular", "salty", or perhaps "a weird taste
I have never encountered before, somewhere between salty and sweet, which
also spills over into a sparkly purple visual sensation".
> re do science on a signal and use the signal to make a
> statement about the natural world that generated/caused the signal
> elsewhere away from/outside the robot. It can't. It's blind, deaf, can't
> taste or smell or touch. Having the sensor transduction does not give it
> sight... that is scientifically proven fact. EG There is a HUGE neural
> sensory transduction/actuation system along the wall of your intestines,
> of which you have no awareness at all, but is hammering away like a
> factory squeezing and pushing all day...
> Put it this way.... a 'red photon' arrives and hits a retina cone and
> isomerises a protein, causing a cascade that results in an action
> potential pulse train. That photon could have come from alpha-centuri,
> bounced off a dog collar or come from a disco light. The receptor has no
> clue. Isomerisation of a protein has nothing to do with 'seeing'. In the
> human the perception (sensation) of a red photon happens in the visual
> cortex as an experience of redness and is 'projected' mentally into the
> phenomenal scene. That way the human can tell where it came from. The
> mystery of how that happens is another story. That it happens and is
> necessary for science is what matters here.
I don't think that's correct. It is impossible for a human to tell where the photon
came from if it makes no sensory difference. That difference may have to involve
other sensations, eg. if the red sensation occurs simultaneously with a loud bang
it may have come from an explosion, while the same red sensation associated with
a 1 KHz tone may have come from a warning beacon.
> The main fact is that the zombie does not have sensation at all and that
> as a result it cannot do science on the world outside the zombie. It's
> doesn't even know there is a world to do science on. All it can do is
> correlate measurements with each other, measurements that could have come
> from anywhere and the zombie can never tell from where.
Alas, I am no better than the zombie in this respect.
Stathis Papaioannou
Be one of the first to try Windows Live Mail.
 You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to
To unsubscribe from this group, send email to
For more options, visit this group at
Received on Sat Nov 25 2006 - 04:22:11 PST

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:12 PST