Re: Kosko on AI in L.A. Times

From: Party of Citizens (
Date: Tue Jul 10 2001 - 21:32:21 MDT

This is very interesting research alright and it will help a lot of
patients. But our interest in AI should cause us to differentiate between
afferent and efferent information flow. The work you describe below is all
efferent flow and that has been demonstrated before at several
centres. Patients are taught for example to move cursors on a computer by
will power. Warwick's imminent experiment will be the first to try out a
major afferent flow of information. Do you acquire instant encyclopedic
knowledge when an encyclopedia on a chip is implanted? That is what
Warwick will find out.

On Mon, 9 Jul 2001 wrote:

> FWP writes:
> > I think we'll learn a lot about chip implants when Professor War-borg does
> > his next experiment early next year. If I understood it correctly, this
> > time he is going to determine if the afferent connections work well enough
> > to upload significant amounts of information directly from chip to
> > brain. Anyay, the article is in Wired, Feb/01.
> That's at
> Coincidentally I received the August issue of Wired today and it describes
> research next to which Warwick's little arm chip is little more than a
> sideshow stunt. The article, by wheelchair-bound John Hockenberry, tells
> about work to expand the capabilities of the disabled via mind-machine
> interface:
> "When you meet Johnny Ray, it's a challenge to see the former drywall
> contractor and amateur musician trapped inside his body, but he's there.
> Ray, a 63-year-old from Carrollton, Georgia, suffered a brain-stem stroke
> in 1997, which produced what doctors call "locked-in syndrome": He has
> virtually no moving parts. Cognitively he's intact, but he can't make
> a motion to deliver that message or any other to the world."
> ...
> "[Dr. Roy] Bakay... wouldn't describe anything he does as show business,
> but to me the results of his work sound like a real-world version of
> the nifty plug Neo/Keanu sported in The Matrix.
> "'We simply make a hole in the skull right above the ear, near the back
> end of the motor cortex, secure our electrodes and other hardware to
> the bone so they don't migrate, and wait for a signal,' Bakay says.
> The implant is an intriguing hybrid of electronics and biology - it
> physically melds with brain tissue.
> "'We use a small piece of glass shaped like two narrow cones into which
> a gold electrical contact has been glued,' Bakay says. 'The space in
> the cones is filled with a special tissue culture, and the whole thing
> is placed inside the motor cortex.' The tissue culture is designed to
> 'attract' brain cells to grow toward the contact. When brain cells meet
> gold, the electrical activity of individual cells is detectable across
> the electrode. Gold wires carry signals back out of the skull, where they
> are amplified. This produces a far more sensitive and usable signal than
> you get from surface technology like the taped-on electrodes used in EEGs.
> "To get a broad sense of what the patient's brain is doing, neurologists
> perform magnetic resonance imaging and compare changes in the motor
> cortext with voltages monitored through the electrodes. Then the doctors
> get really clever. The patient is encouraged to think simple thoughts
> that correspond to distinct conditions and movements, like hot/cold or
> up/down. Gradually, the doctors extract and codify electrical patterns
> that change as a patient's thoughts change. If a patient can reproduce
> and trigger the signal using the same thought patterns, that signal can
> be identified and used to control, say, a cursor on a computer screen.
> The technique is very crude, but what Bakay and his colleagues have
> demonstrated is a truly alternative brain-body interface pattern."
> At first Ray has to imagine arm motions to trigger the cursor, but over
> time his brain adapts:
> "As the brain cells in and around Ray's implant did what he asked
> them to do, the imagined sensation of moving his body parts gradually
> disappeared altogether. One day when his skill at moving the cursor
> seemed particularly adept, the doctors asked Ray what he was feeling.
> Slowly, he typed, 'nothing.'
> "Ray was interacting directly with the cursor similar to how he might once
> have interacted with his hand. 'People don't think "move hand" to move
> their hands unless they are small children just learning,' Bakay explains.
> 'Eventually the brain just eliminates these intermediate steps until the
> hand feels like a part of the brain.' The description reminds me of how
> I've heard Isaac Stern describe his violin as an extension of his body.
> I think of my wheelchair the same way.
> "The fact that Ray's cursor is indistinguishable from almost any other
> prosthesis raises an important philosophical question: Because of the
> implant, is a Dell Pentium cursor now more a part of Johnny Ray than
> one of his own paralyzed arms?"
> These people are the real cyborgs, with actual brain implants interfacing
> to their nervous system, allowing them to control the world around them
> in new ways.
> Hal

This archive was generated by hypermail 2b30 : Fri Oct 12 2001 - 14:39:43 MDT