Connectionists: How the brain works

[Geoffrey Hinton]

Actually, evolution did invent the time-shared wheel.

To go over rough ground it needs to be 6 feet in diameter with very soft
suspension.
The way to do this without being too heavy or large is to time-share two
small sections of the rim each connected to the axle by "spokes" (in
compression rather than tension) that can easily change their length.  The
swapping in and out of the spokes is not as energy efficient as a wheel but
it solves the problem of supplying the rim with nutrients.

Geoff



On Mon, Jan 27, 2014 at 9:28 AM, Balázs Kégl <balazskegl at gmail.com> wrote:

> > While it is at least worth considering whether the arm from fin argument
> applies to the nervous system, because we don’t understand how the brain
> works, we can’t really answer the question whether there is some simpler
> version that would have worked just as well.  Accordingly, as with the
> radio analogy, in principle, asking whether a simpler version would work as
> well, depends on first figuring it out how the actual system works.  As I
> have said, abstract models are less likely to be helpful there, because
> they don’t directly address the components.
>
> Wouldn’t the airplane/bird analogy work here? Does being able to design an
> airplane help understanding how birds fly? I think it does. Evolution
> didn’t invent the wheel, so it had to go in a complex (and not necessarily
> very efficient) way to “design” locomotion, which means that airplane
> engines don’t really explain how birds propel themselves. On the other
> hand, both have wings, and controlling the flying devices looks pretty
> similar in the two cases. In the same way, if some artificial network can
> reproduce intelligent traits, we might be able to guide what we’re looking
> for in the brain (a model, whose necessity we agree on). Of course,
> scientific process rarely works in this way, but it’s because you need
> computers for this kind of “experimentation”, and computers are quite new.
>
> Balázs
>
>
> —
> Balazs Kegl
> Research Scientist (DR2)
> Linear Accelerator Laboratory
> CNRS / University of Paris Sud
> http://users.web.lal.in2p3.fr/kegl
>
>
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20140127/1da9ed5c/attachment.html>