No, the idea being that thinking that a hard drive will allow you to store things other than what is exactly designed to be stored on a hard drive completely misunderstands the technology and the underlying physical principles.
No, the idea being that thinking that a hard drive will allow you to store things other than what is exactly designed to be stored on a hard drive completely misunderstands the technology and the underlying physical principles.
Vitiate Man.
History repeats the old conceits
The glib replies, the same defeats
Spoiler Alert, click show to read:
I'm not sure though whether the idea that artificial neural networks could one day be as good as biological ones is that crazy.
What sounds like a lack of understanding to me is calling them "neural networks" and "actual neurons", because to my understanding, "actual neurons" only work really well in a "neural network". If that's what you mean then I understand you, but your quote is also a bit off given that the quoted person is not the one making the mistake there.![]()
![]()
![]()
"Topic is tired and needs a nap." - Tosa Inu
If neurons act as a neural network, then a computer software designed for certain tasks and using select characteristics of some neurons will not behave as a "real" neural network.
Lightbulb technology will never be "as good as" a sun when it comes to generating light, both because that's an inherent technological constraint and because the design of a lightbulb, being made for specific purposes, is very far from the design of a star.
Vitiate Man.
History repeats the old conceits
The glib replies, the same defeats
Spoiler Alert, click show to read:
But now you entered the constraints of using select characteristics and being designed for specific tasks.
My point was that these things are still in development, it seems strange to assume that they cannot one day do more.
And it doesn't just have to be software, apparently some chips are designed as neural networks: http://www.research.ibm.com/cognitiv...ic-chips.shtml
That is a very strange way to combine these arguments. The physical limitations are relatively obvious, but to say a light bulb is created for specific purposes that differ from those of a star is a bit odd given that the purpose of a light bulb is to replace the light of a star in places or at time where the light of a star is not readily available. Of course the design is different, having a star with the smallest working mass on our planet would be a bit much and also a bit hote, in fact one would rather say having our planet on such a star rather than the other way around.
The light bulb still allows you to see things the same way the star does, so with that purpose in mind, they both provide the same result.
If you are refering to the idea that pouring alcohol over a brain vs pouring it over a computer chip on which a neural network is running yields different results, well, whom are you quoting, Captain Obvious?![]()
![]()
![]()
"Topic is tired and needs a nap." - Tosa Inu
What they will be able to do in the future depends on how are present assumptions and what we think we are trying to accomplish are revised. The important thing is that 'more of' what we're doing now will not push boundaries.it seems strange to assume that they cannot one day do more.
And that's exactly what current approaches can manage - and that's fine. But let's not pretend we're building stars when we're just making plain-old lightbulbs.they both provide the same result.
Vitiate Man.
History repeats the old conceits
The glib replies, the same defeats
Spoiler Alert, click show to read:
Is a promotion without a pay raise (or any pay at all) really still a promotion?
EDIT: not talking about my recent donning of the red robes.![]()
On the Path to the Streets of Gold: a Suebi AAR
Visited:![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
![]()
Hvil i fred HoreToreA man who casts no shadow has no soul.
No, it's spot-on. Echoed here.That's fine, but the comparison with astrology sounds a lot meaner than that.
Vitiate Man.
History repeats the old conceits
The glib replies, the same defeats
Spoiler Alert, click show to read:
Guys, please look up the facts before making statements like these. Wikipedia gives a solar luminosity figure of 98 lumens per watt. I can buy a Cree LED 100W equivalent light bulb that out puts 1,650 lumens at 16.5 Watts. We are already there and the technology is still improving.
First half of your statement was that lightbulb technology will never be "as good as" a sun when it comes to generating light because of "inherent technological constraint".
My point is that we already have lightbulbs that are better at generating light than the sun.
Why do you say that? Your citing of the metric offers no context to what we make of the sun - and note that with respect to the sun the wattage component of the efficacy apparently refers to the radiated heat rather than direct source electrical consumption or conversion.My point is that we already have lightbulbs that are better at generating light than the sun.
No, what you have cited goes toward evaluating some lamps against others in generation of light. Don't misuse metrics in response to a criticism of the misuse of metrics.
Vitiate Man.
History repeats the old conceits
The glib replies, the same defeats
Spoiler Alert, click show to read:
Bookmarks