Upgrade your account to a Community Supporter account and remove most of the site ads.

Reply to thread

Message: <blockquote data-quote="Blue" data-source="post: 9374823" data-attributes="member: 20564">Yes.  I started with a discussion of us not knowing about the brain, but then talked about the conscious process we take.  You've spent a good number of words not actually addressing what I wrote that are moot for this discussion.No, AI art does not.  Flat, period, not up to discussion.  AI art is not generative each item in 3d, determining where it is, what's visible, what light is doing, and the like.As mentioned, there are definitely products that duplicate how we do this, but it's not what's being trained on the art out there.  Nor are they "creative".You were okay until the last sentence.  If you want to assert that our brain is a statistical model, you need some supporting evidence.Which is relevant to the topic at hand how?  We're discussing if AI model training and the produce an output is the same as how humans do.  Sentience is a whole different ball-field.In the same discussion where I called it "spicy autocorrect", I talked about the huge number of iterations, from horse and buggy to a top of the line sports car.  Don't try to dismiss the point via rhetoric.Also, NO you can't call human language production spicy autocorrect because it doesn't work the same way as .  Which is the point at hand.  That is absolutely not a given.  Don't confuse your assumptions for truth.The first L in LLM is also something to look at, it stands for "Large".  The amount of input data needed to generate those statisical data is far, far beyond what a human being is exposed to in order to generate speach.  Many orders of magnitude.  Because humans are not using the same method.Nope.  Not in the slightest.  Strike that all important word "statistically" and replace it.  I'm not sure with what, maybe "conceptually" but I'm not married to that.  Studies on single vs. multiple language households are interesting, especially when grammar is different so "statistically" there are very different word orders for what would come next but there can, among humans, still concept replacements in one language with grammar of the language a sentence has been started in when a word for a concept is only know in the other language.Nope, didn't bring it up.  A completely different conversation then the one I've been having.  I bring up concepts because it's how humans do it.  Which is different than the statistical analysis from the LLMs.  I'm not attributing concepts to conciousness.  I am saying that the method used under the covers differs, since the brain is not doing statistical analysis as the primary or solo method either to generate art nor conversation.</blockquote>

Verification