Re: Artificial life likely in 3 to 10 years

DK wrote:
Tim Tyler <seemysig@xxxxxxxxxxxxxx> wrote:

Moravec bases his estimate on the functional properties of
retinal cells. There is a sketch of the calculation here:

``Each of the million ganglion-cell axons carries signals from
a particular patch of image, representing the differences in
light intensity between adjacent regions and from one time to
the next--edge and motion detections that are useful also in
robot vision. Overall, the retina seems to resolve about one
million distinct regions in the visual field and to follow
change up to about ten frames per second. Fed a video image
with similar resolution, it takes a robot vision program about
100 computer operations to produce a single edge or motion
detection, thus 100 million operations to match a whole
"frame" of optic nerve output, and 1,000 MIPS--the power
of a small supercomputer--to equal the retina's ten frames
per second.

If the retina is worth 1,000 MIPS, what about the whole brain,
whose larger neurons are 1,000 times as numerous, but occupy
100,000 times the volume? Multiplying the retina's computation
by a compromise brain/retina ratio of 10,000 yields a rough
brain equivalent of 10 million MIPS--like a million 1997
robot computers, or 100 of the biggest supercomputers. [...]

The author obviously never observed small fish behavior.
That's the problem with some computer scientists - they live in
a bubble where it is easier to get some computer numbers than
to simply observe aquarium fish and realize how wrong their
numbers are.

Had Moravec ever observed an advanced experimental robots
that have personality? Fish do have it.

That is a ridiculous argument :-(

You might as well argue that computers can play chess and fish
can't therefore, computers are smarter than fish. The logic of
that is of about the same level.

http://www.frc.ri.cmu.edu/~hpm/project.archive/general.articles/1997/970128.nos
ense.html

Of course, this is just an estimate.

Of course it is. And a wrong one. Brain does not process
information linearly. In the above example retina is just a video
camera.

The retina does perform signal processing - thus the reference
to edge and motion detection.

If Moore's law continues
to hold, for each factor of two is is out by, you need to add or
subtract 18 months. Say, if you think it is out by a factor
of 1000, that would correspond to a difference of about 15 years.

Oh, since the article is 10 years old, then we are almost
there?

Moravec was doing the same calculation 20 years ago - e.g.:

``Average humans beings can be beaten at arithmetic by a one
operation per second machine, in logic problems by 100
operations per second, at chess by 10,000 operations per second,
in some narrow "expert systems" areas by a million operations.
Robotic performance can not yet provide this same standard of
comparison, but a calculation based on retinal processes and their
computer visual equivalents suggests that a billion (10^9)
operations per second are required to do the job of the retina,
and 10 trillion (10^13) to match the bulk of the human brain.''

http://www.frc.ri.cmu.edu/~hpm/project.archive/general.articles/1988/takeover.ltx

His timescale then was:

``I will argue that robots with human intelligence will
be common within fifty years.''

Computer Go, provides a better yardstick, IMO:

"To Test a Powerful Computer, Play an Ancient Game"

- http://www.ishipress.com/times-go.htm

FWIW, my predictions for progress in computer go are
shodan: 2020, champion: 2050.

You continue to talk about very complex calculators.

"Calculator" was not a term I introduced into this discussion.

And the whole thing is that brain is not a calculator!

Functionally, it's a computing device. Its main role
is to accept inputs, process them and produce outputs.

Instead, every single cell is a supercomputer and a brain
is a self-regulating P2P network consisting of 10^11 of
these computers.

Neurons are certainly not analogous to supercomputers
in their functional role in the brain, as signal
processing devices. It's more down to the fact that
there's a hundred billion of them that gives the
brain its processing power.
--
__________
|im |yler http://timtyler.org/ tim@xxxxxxxxxxx Remove lock to reply.

.

Relevant Pages

  • genes and sight
    ... Gene Therapy Awakens The Brain Despite Blindness From Birth ... demonstrated that gene therapy used to restore retinal activity to the ... "The retina of the eye captures light, but the brain is where vision ...
    (uk.philosophy.humanism)
  • Re: AInet: HAL 9000 wants YOU!
    ... packets to take in the human brain is mimickable by manmade systems. ... But he equates 10,000 MIPS to guppies, 20,000 MIPS to lizards, 100,000 ... computation power it would take is mute until someone figures out how ... The retina pre-processes detections of edges and/or motions ...
    (comp.ai.philosophy)
  • Re: F/0.5 possible?
    ... The retina itself does an amazing amount ... Many vision physiologists do indeed say the retina is part ... of brain and does neural processing. ... Physiologists are doing some very interesting work on the mid-brain ...
    (rec.photo.digital)
  • Re: Retreat From Reason
    ... of view of the coloured image that appears on the retina. ... particularly for colour vision. ... if you claim human eyes are like Zeiss cameras ... the brain or is at least an outgrowth of the brain. ...
    (sci.med.vision)