Talk:On Intelligence
From Wikipedia, the free encyclopedia
Contents |
[edit] Informal Dendrite Theory: Opinions?
Although we do not yet know the precise chemical/electrical sensitivities of the 'long range' neuron sensor/conduits called dendrites, this should be no cause for concern. Functionally their purpose is made clear in On Intelligence and hypothesizing should be a matter of 'filling in the blanks'. Of course it is helpful to enhance their possible role with hypothesis, rather than trying to 'sweep them under the carpet' with downplay. A positive approach to this is simply to ask, 'What is the advantage of a dendrite?' or 'What can it do, what role(s) might it play, in boulstering such an interconnected symphony of modulated signals?' At least, that's how I would put it.
Dendrites might function as 'scouts' for major neurons, located at considerable distance and maintaining subtle, but meaningful bonds with the 'major conduits'. Some are quick to forget that only small areas of columns are 'lit up' at once, while most of the cortical mass is on 'stand-by'. And although signal travelling 'up' from the powerful senses, or down from the external thalamus, may be tremendous during 'rush hour', is there not a breeze before the storm? In otherwords, although lateral dendrites may not be the 'stars of the show', their role may be to presage or echo large events in the cortical mass. This would describe a transition period, a 'soft weather', allowing for greater fluidity and preparedness in violent lateral shifts of up/down spike matching.
But this too is focusing on very high stages of neural activity, easily recorded by our primitive scanning techniques. Can we really measure the full subtlety of dendrite actuation? Although major neurons clearly play a 'highway' role in transmitting and sustaining pulses, are they (in their admitted ability) the whole story? Or are we not being counterintuitive enough...? Does such a major cell exist without long range sensitivity? Indeed, is it not vital in order to predict distantly adjacent action potential, to have scouts, watchdogs, at the ready in all directions? They would certainly be either a) inhibited during 'more important matters' or, b) simply inaudible/inconsequential during major neuron pulses. But as Mr. Hawkins made very clear, coincidence, subtlety, smartness and connectedness would be the dendrite advantage. They would be able to read very subtle shifts in electrical potential, the wind of the mind, pointing the direction to the greatest potential chaos. It may never emerge, but if it is either a) neighboring close enough or b) far away, but really really powerful, would dendrites not be ideal sentrys? And used in unison, focusing, yes, on coincidence, many of these single dendrites sent out from a neuron could inform the cell as to wether there was 'regular white noise' or 'maybe a seizure'.
Athough most of the brain is considered sedate at a given moment, our scanners cannot rule out more subtle shifts in probability brewing slowly in our cortex. And with such a charged piece of biomass constantly at the ready, shouldn't we start to give smaller participants a little more credit? The memory-prediction framework seems to leave open the potential for 'unpredictability' when the brain encounters 'special circumstances.' A great deal of pressure seems to fall on the individual cell, which has to make a choice informed on something. Lateral connectivity, although not certain, could perhaps lie in the shifting of the winds.
[edit] Deer in the headlights
Hawkins' prediction: "exception cells" should remain OFF during a learned sequence obviously does not apply to to a deer in the headlights of a car. In this case, the deer's brain is mesmerized, and full of alarms about the car, and it will take some other stimulus to startle the deer out of its trance so it can run to safety. Ancheta Wis 23:18, 18 September 2005 (UTC)
I added an image. Did I do this correctly? What about copyright issues? Atraxani 01:52, 14 November 2005 (UTC)
[edit] Grandmother neurons
Here are some Google hits for Hawkins' "Name cell", or "grandmother cell"; additionally, note a "Bill Clinton neuron" discovered in a subject's hippocampus by Christof Koch; other cells in that subject's hippocampus were the "Jennifer Aniston neuron" and the "Halle Berry neuron", etc. 20:22, 8 April 2006 (UTC)
- A related concept mirror neuron has been developed by other researchers. In primates, mirror neurons fire (discharge) when an action known to the subject is performed by others; 'monkey see monkey do' is enabled by the existence of mirror neurons, which allow us and other primates to imitate each other. The concept was discovered when a primate observed a human pull a banana from a bowl, and a mirror neuron fired.[1] --00:43, 30 October 2006 (UTC)
- ^ Scientific American special issue on the Senses 2006
[edit] I miss a criticism section
Hi there. I'm reading the book (which i believe falls in the category of popular science) while having my first psychology-lectures at university.. does it deserve a criticsm section ? are there any voices about that ? thx in advance. —Preceding unsigned comment added by 85.178.145.235 (talk) 13:59, 16 October 2007 (UTC)
