Discussion Topic |
|
This thread has been locked |
Largo
Sport climber
The Big Wide Open Face
|
|
|
Topic Author's Reply - Jun 26, 2018 - 05:56pm PT
|
What I think is that both of you really don't get how more complicated things can be built from fundamental building blocks along with some processes and rules.
This is the complexity argument, re: a super complex and complicated combination of objective functions "explains" mind, when in fact there is no reason whatsoever why complexity should "cause" any experiential quotient. None.
In virtually ever instance listed on this thread, putative "explanations" are in fact descrioptions of WHAT happens, arrived at through direcxt observation. NOT reasons why a given force or phenomenon arises as opposed to some other force or phenomenon - or nothing at all.
|
|
paul roehl
Boulder climber
california
|
|
|
Jun 26, 2018 - 06:10pm PT
|
I have no need for a definition of intention.
The problem is an inviolable order implies intention. Call it the intention of the universe or the intention of god but there are certain things matter and energy do and certain things they do not do. The universe is something we seem to be able to intelligently understand precisely because of that order and the implication of our understanding of the continuum of that intelligence implies an intelligence beyond our own. God may or may not be, but one can't ignore or really dismiss the serious reality of these implications.
|
|
WBraun
climber
|
|
|
Jun 26, 2018 - 06:31pm PT
|
And they're getting smarter everyday.
Smart is not the same as intelligence.
Just like you are so called smart but lack good intelligence ......
|
|
yanqui
climber
Balcarce, Argentina
|
|
|
Jun 26, 2018 - 07:20pm PT
|
Meanwhile, AI bots just beat humans big time in said game. And they're getting smarter every day. Sheesh!
Yet robots still can't walk around a room and recognize the objects that a 3-year old can.
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Jun 26, 2018 - 08:30pm PT
|
...when in fact there is no reason whatsoever why complexity should "cause" any experiential quotient. None.
well I don't think that is entirely true, and many here argue just the opposite when they invoke human exceptionalism, that is, humans are more "complex" than other organisms.
If you are going to make an argument based on complexity, then you should be ready to explain what complexity is, and that turns out to be more difficult than it would appear at first blush.
For information, as I alluded to above, Kolmogorov complexity works well. For some output, say a string of text, its Kolmogorov complexity is determined by the computational resources required to produce it, e.g. the length of the code.
The example of mb1 above of the SSL certificate which can be generated by a number of algorithms, RSA being one. The pseudocode would be something like:
1) choose two very large (e.g. 1024 bit) random prime integers: p and q
2) compute n=pq and phi(n)=(p-1)(q-1)
3) choose an integer e between 1 and phi(n) such that the greatest common denominator of e and phi(n) is 1
4) compute d between 1 and phi(n) such that ed= 1 (mod(phi(n))
the public key is (n,e) and the private key is (n,d)
the values of p, q, and phi(n) are private
e is the public (encryption) component
d is the private (decryption) component
so not very resource intensive (which is why it is so useful), and thus not complex.
from this stand point one might look at genetic material as the "body plan" for organisms, and estimate organism complexity based on the genetic resources required to generate the "body" in analogy to the Kolmogorov complexity.
We get to the somewhat surprising result that there is not a great variation in the complexity of living organisms due to our shared genetic heritage.
Actually a delightful speculation.
|
|
WBraun
climber
|
|
|
Jun 26, 2018 - 09:16pm PT
|
So the gross materialists only analyze the different coats (material bodies) of the living entities and completely miss the living entities themselves.
Thus they come to another incomplete faulty dead end .....
|
|
MH2
Boulder climber
Andy Cairns
|
|
|
Jun 26, 2018 - 09:42pm PT
|
Yet robots still can't walk around a room and recognize the objects that a 3-year old can.
If we had a metric, or just a yardstick, for the difficulty of such problems. Maybe a 3-year old is not a bad choice, but how good at chess is the best 3-year old? Maybe chess just isn't that hard?
And how do we compare apples with oranges?
edit:
And I agree with you, as does R Llinás
http://www.supertopo.com/climbing/thread.php?topic_id=1593650&tn=244
|
|
MH2
Boulder climber
Andy Cairns
|
|
|
Jun 26, 2018 - 10:22pm PT
|
NOBODY makes a strong case for "evolution" in the sense of adaptation producing reproductive isolation
But how about what happens when geology, for example the uplift of the Himalaya or Andes mountains or the Tibetan plateau, or the Isthmus of Panama (for marine life) produces reproductive isolation.
Adaptation probably depends on mutations and their suitability (almost always poor because of their random nature) for a particular environment. The environment can also change because of geology, such as continental drift or uplift or subsidence.
|
|
High Fructose Corn Spirit
Gym climber
|
|
|
Jun 27, 2018 - 08:15am PT
|
eeyonkee, imo, you need to back off this thread for awhile. My recommendation: Go to Steam and play DotA 2 for a few days to few weeks. Take some time. Decompress. :) -hfcs
Oh, dear. Eeyonkee, I hope you - any (casual) lurkers as well - didn't take this as any form of criticism. It was not. I meant it totally humorously. (After just reading up about OpenAI's AI's advances in this game.)
I think between you and me, you know I appreciate your postings.
...
Re definitions: If memory serves either David Deutsch or Mag Tegmark (?) on the Waking Up podcasts (last year?) gave remarkably lucid definitions for "knowledge" and "computation" and "intelligence" and "science", etc. - and even "person" for use in AI and game theory/ gamification, etc.
Maybe I'll try to find later. In the meantime, my notes from listening to this podcast say... re "knowledge"
"Philosophers have almost defined it out of existence." -Deutsch
lol
https://youtu.be/J21QuHrIqXg?t=5m30s
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Jun 27, 2018 - 08:29am PT
|
for those who like to read (rather than view video)
What is Computation? (How) Does Nature Compute?
David Deutsch
"The latter is the very misconception that led Zeno astray, and Hilbert and Kant, and many other thinkers throughout history, who haven’t realized that while truth can be absolutely necessary and transcendent, all knowledge (even of such truths) is generated, computed, by physical processes, and the scope and limitations of such knowledge are conditioned by Nature’s contingent laws."
|
|
MikeL
Social climber
Southern Arizona
|
|
|
Jun 27, 2018 - 10:07am PT
|
eeyonkee: Please tell us Mike what you mean by "seem to be"? It does not surprise me at all that you would say this, I just want to know why.
(I wrote that MB1’s criticisms about scientific claims regarding the standard model of evolution theory seemed reasonable and significant to me. I was of the same opinion with the little that I’ve read.)
eeyonkee,
I would say that everyone comes to his or her own conclusions about things—whether to accept authority; whether to go along with consensus; whether to trust their own perceptions, thoughts, or feelings; whether reason or mathematics “proves” anything or is superior to other means by which to generate understanding; etc.. You can see exactly that sometimes from DMT: there are certain things he will not be moved on. It’s not intransigence; it’s perhaps more a feeling of certainty that he’s at peace with.
What MB1 has done a good job of, imo, is exposing the underlying foundations by which to answer any of the questions that I’ve written in the above paragraph.
Socrates was a guy who seemed to be really good at bringing people to direct apprehensions rather than by the understanding of logic, words, etc. The result of which is referred to as “noesis.” Noesis means “just getting it.”
In the end, everything is noesis. All methods lead to noesis.
As moderns, we know things through rational processes, but we cannot reason without having an idea of what counts as truth. For that, we must consult a vision. (What’s your vision?)
It’s a paradox, you know. To know what’s reasonable we must know what counts as the truth to know if we’ve reasoned properly or correctly. So too with the Good and the Beautiful I referred to in the earlier post. All 3 supposedly are looking in their own ways at the same thing. We must know these things (what counts as The True, The Good, and The Beautiful) before we can know anything else otherwise we can’t reason at all. These things (True, Good, Beautiful) we know differently than we know anything else (via typical disciplinary or specialized knowledge taught in schools).
All 3 pillars are just names and powerless without experiences to ground them. What is important, is the vision itself. Everything else is secondary. To Largo’s point above, explanation is not the point; it is only the pointer. “Once you’ve climbed the ladder, you can kick it away” (Wittgenstein).
You said you were going to make some claims without reading or referencing public authority (Wiki). You feel truth in your bones—due to your education, readings, work life, etc. That’s noesis.
Be well.
|
|
eeyonkee
Trad climber
Golden, CO
|
|
|
Jun 27, 2018 - 10:33am PT
|
Sorry for being rude, Mike. Okay, so now I am looking up the term and here is the first thing that I came up with -- from Merriam-Webster. Note 2b and 2c -- pretty much what I have been saying, and contrasts with what MB1 has been saying. In fact, 2b conforms nicely with the thought experiment that I asked MB1 about.
Definition of information
1 : the communication or reception of knowledge or intelligence
2 a (1) : knowledge obtained from investigation, study, or instruction (2) : intelligence, news (3) : facts, data
b : the attribute inherent in and communicated by one of two or more alternative sequences or arrangements of something (such as nucleotides in DNA or binary digits in a computer program) that produce specific effects
c (1) : a signal or character (as in a communication system or computer) representing data (2) : something (such as a message, experimental data, or a picture) which justifies change in a construct (such as a plan or theory) that represents physical or mental experience or another construct
d : a quantitative measure of the content of information; specifically : a numerical quantity that measures the uncertainty in the outcome of an experiment to be performed
|
|
jogill
climber
Colorado
|
|
|
Jun 27, 2018 - 11:07am PT
|
"Philosophers have almost defined it out of existence." -Deutsch
Not bad. Probably more time and effort might productively be spent on content. Word games. Almost as trying as "being."
|
|
eeyonkee
Trad climber
Golden, CO
|
|
|
Jun 27, 2018 - 02:41pm PT
|
"The latter is the very misconception that led Zeno astray, and Hilbert and Kant, and many other thinkers throughout history, who haven’t realized that while truth can be absolutely necessary and transcendent, all knowledge (even of such truths) is generated, computed, by physical processes, and the scope and limitations of such knowledge are conditioned by Nature’s contingent laws."
Can't imagine why you like this one, Ed:) The more I think about it, the more it resonates. Remember how I was saying a while back that I strive to express my position with brevity? Well, I'm stealing this one.
Nature’s contingent laws -- Sheesh, why hadn't I thought of that phraseology? It's beautiful! Hard to get below three words.
Bonus FYI. Here are concepts in computer programming that align closely with the biological world.
 Contingency
Hierarchy/dependency
Composition
Recursion
Life history
Replication
|
|
jogill
climber
Colorado
|
|
|
Jun 27, 2018 - 04:03pm PT
|
"Remember how I was saying a while back that I strive to express my position with brevity?"
Brevity is the soul of wisdom. Be concise. Be accurate. Get results. Avoid rambling and get to the point. One reason classical mathematical analysis is appealing. (Not so the modern abstract variety, IMHO - but, I'm old and worn down)
|
|
MikeL
Social climber
Southern Arizona
|
|
|
Jun 27, 2018 - 04:08pm PT
|
:-)
I didn't think you were rude, eeyonkee.
It's hard to think of any question that could be said to be rude.
|
|
WBraun
climber
|
|
|
Jun 27, 2018 - 04:25pm PT
|
all knowledge (even of such truths) is generated, computed, by physical processes, and the scope and limitations of such knowledge are conditioned by Nature’s contingent laws."
Not true ever ....
The inferior Material Nature (energy) is always subordinate to the Superior Spiritual Energy.
Thus "all knowledge" originates NOT from Material Nature but from the Spiritual stratum itself.
|
|
madbolter1
Big Wall climber
Denver, CO
|
|
|
Jun 27, 2018 - 09:35pm PT
|
The latter is the very misconception that led Zeno astray, and Hilbert and Kant, and many other thinkers throughout history, who haven’t realized that while truth can be absolutely necessary and transcendent, all knowledge (even of such truths) is generated, computed, by physical processes, and the scope and limitations of such knowledge are conditioned by Nature’s contingent laws.
I've been buried, and I come back to find that there is simply no way to even begin to address everything. So, I'll focus on two points:
1) Nothing in my argumentation has appealed to authority. My arguments stand or fall on their own. It's a pretty ridiculous mistake to descend into groupthink, believe that consensus is how metaphysical facts are determined, or that "authority" makes any difference in this discussion. Arguments stand on their own to be evaluated for their basic logic and coherence with the objective facts. If, for example, I say, "There are no genuine examples of ring species," there's no appeal to authority or consensus. It's trivially easy for anybody to look at the few putative examples of ring species and see that the lead researchers in each of those cases later repudiated their own earlier conclusions. There are just facts of the matter and arguments based upon them. Evaluate for yourselves and come to your own conclusions on both the facts and the arguments.
2) The above quoted passage is so ridiculous that I literally cannot do justice to all of the things wrong with it. I can fairly quickly say just a couple of things.
A) The supposed profound insight is: "Kant, and many other thinkers throughout history, who haven't realized...." And then it goes on to outrageously and even laughably make Kant (and many other thinkers) to be saying something they didn't say or believe; in fact they flatly and obviously articulate the VERY "profound" insight this passage states, because it's OBVIOUS to the point of being almost trivially true.
OF COURSE there's a difference between truth (as in facts of the matter) and knowing; that's precisely the metaphysical/epistemological divide! What does this goofball (I say that intentionally) think that Kant's project even WAS??? Duhhh, I mean, seriously, I cannot take a "thinker" seriously who pops off with such tripe as though it is all wise and profound, when you don't have to know hardly ANY Kant to know that that assessment is transparently ridiculous! NO serious philosophical critique of Kant is based on this tripe.
B) The take-away is: "...the scope and limitations of such knowledge are conditioned by Nature's contingent laws." Well, hey, you can CLAIM it all you want; such CLAIMS don't make it so. This is just mapping the empiricist perspective onto reality rather than seeing if reality can be accounted for on strictly empiricist terms.
But, as I said upthread, there's no "end game" here. Empiricists MUST have something like the quoted perspective, and I believe that NOTHING can be done to broaden their view. Empiricism is a zero-sum game, and empiricists cannot get out of that box. So, my goal is not to "convince" or "convert" empiricists. Can't be done.
As I said, my goal has been a pretty low bar, and I believe that I'm over it to my own satisfaction: Make the case that the sweeping dismissals of a non-empiricist view are not necessarily issued by all and only stupid, ignorant, or wicked (a la Dawkins) idiots. There is NO way in threads like these to write chapter-length, systematic rebuttals of ALL the dust that empiricists can keep hanging in the air. ALL I've sought to do is focus on a few particles and say, "Hey, not so fast." And that's literally about the only extent to which I can address all this stuff in this venue.
The idea that human thinking can all be reduced to some "computational" model is a non-starter. Humans are capable of thinking in ways that computers will never do, sci-fi not withstanding; the promise of hard-AI is a logical implication of this reduction-to-computational model. However, it will never happen; that's the prediction of a non-empiricist.
Everything computers do and in principle can do is deterministic. Even if the promise of quantum computing is fully realized, computer programs cannot in principle "get outside their box," so everything they "think" can ultimately be traced back to some particular coding. That is not the case with human thinking, and that's not "just" because we can't yet "get at the code."
What humans can do that computers in principle can't is adopt a "meta" perspective on any subject. Just one example is in the field of polyadic predicate logic with identity. It is a non-deterministic, non-guaranteed-outcome process to produce both derivations and invalidating models for such theorems and arguments. Humans can reliably do it, but computers dismally fail at it. Some programs have been designed that can "take stabs at it," but their outcomes can be directly tied back to the sorts of "strategies" that are programmed in, and it's straightforward to "respond to the known strategy" by throwing a derivation at the computer that doesn't fit what it knows or CAN know from the known coding.
The qualitative difference is that humans can reliably do it on any arbitrary theorem or argument. And anybody that's studied this stuff knows that doing such derivations or finding invalidating models is, shall we say, non-trivial! Yet humans with pretty minimal training can reliably do it and on any range of theorems or arguments. Minimal training simply teaches the syntax and rules, and from there, humans appear to be immediately unlimited in their capacity for this stuff.
Moreover, humans can move from doing that sort of stuff to asking the "meta" questions about that very stuff! You can't get computers to do the former, MUCH less the latter.
The uninitiated think something like, "Well, Big Blue can kick azz in chess, which is also 'non-deterministic' and offers a vast range of combinations." NOT true. I don't know Big Blue's particular algorithms, but I know that they are deterministic and that Big Blue CANNOT "step back" and ask the meta question, "What's the point to chess?" Or, even more meta, "What's the point to asking 'what's the point?'" Big Blue functions entirely within its limited-scope context, and asking ANY meta questions are beyond its programming.
The uninitiated then say, "Well, Big Blue is a 'learning system,' so it certainly is asking SOME 'meta' questions, or it could not learn." And there the smuggling begins.
Big Blue does not LEARN in anything like the sense that humans do. Big Blue is not asking a 'meta' question like, "Why didn't that strategy work as well as hoped?" Big Blue computationally "evaluates" outcomes and keeps track of statistics. It does so within deterministic rules. And there WILL ultimately be some human chess player that will tumble to HOW Big Blue processes, extrapolate the deterministic rules that guide it (including the ones it has 'learned') and then start systematically beating it, even given its vastly greater resources devoted to the ONE problem it is faced with.
Reprogram Big Blue with the rules of polyadic predicate logic with identity derivations, and let's compare its success rate with that of a human being then. Then, let's see Big Blue do anything approaching the meta perspective achieved by, say, Godel. Not gonna happen. Ever.
Computers don't have INSIGHTS. They do not "think outside the box," nor CAN they, and quantum computing isn't going to cause such a qualitative change to occur.
Those are my predictions for the promise of hard-AI, and I do AI for a living.
I know in what sophisticated ways you can SPOOF thinking in machines, and I know HOW that spoofing is happening. The Turing Test should not be merely about whether a machine can "fake out" a person. It should revolve around computers PROVING that they can "think outside the box" in ways that cannot be deterministically traced back to their programming. Let them start generating genuine INSIGHTS that have humans struggling to understand at all with implications beyond what any human has ever thought before.
That is not going to happen, yet humans do this sort of thing all the time.
Let a computer EVER produce a genuine "meta" thought that demonstrably was not just a direct result of its programming (I mean, that you can replicate such 'thoughts' from particular code-snippets that directly have the implication of such a 'thought'), and I'll retract everything I'm saying about the reduction of knowledge to computation.
At present and forever, computational "knowledge" amounts to experiments in spoofing, like, "How cleverly can we make what we KNOW to be entirely deterministic 'thinking' PRETEND that it's acting non-deterministically?" And the most impressive results you get out of such computer-"thoughts" is something like, "Oh, that's an implication I hadn't immediately realized, but yeah, with those inputs, that actually is the result." That's not INSIGHT or "meta" thinking. That's just implication and deterministically so. That's no Godel, or Einstein, or Beethoven, or....
And when computer AI screws up (as I know first-hand), you can dig into the code and ultimately discover how that "edge case" implication emerged, which amounts, literally, to the computer doing exactly what you told it to do but with an implication from that that you did not calculate for. It's not "meta!" It's a MISTAKE in YOUR coding, and it doesn't lead to INSIGHT! It leads to MISTAKES.
In our system, a mistake in our AI coding NEVER has us thinking, "Oh, wow. That degree requirement actually DOES mean something completely different from what we thought." It is always and necessarily a MISTAKEN implication where we left an edge case undetermined, and the computer chose a default path we hadn't foreseen (or just had the audit blow up entirely) that was just WRONG. You can only SPOOF "undetermined" via deterministic methods. But human beings think the undetermined and the genuinely "meta". That's a qualitative divide that computation will not cross.
Anyway, I could spend the rest of my life in such arguments. On this thread at least, I tire of it. There IS no "end game," and the arguments rage endlessly and always will. I know that I'm in the tiny minority, and, just as at this present moment, I cannot possibly respond to all the "moles" that need a good whacking. Hehe
It's not my job, man. LOL So, it's only good as long as it's fun. But it's too much work at present to remain fun.
It's been fun thus far, but now it's feeling more tiresome than fun. Call that a "win" if you please. I don't need to "win" or leave feeling like I've "won" anything. But arguing at this sort of level is difficult and time-consuming, and I do have paying work (and some recreation) to do. So, I might drop in now and then, but I have no need nor ongoing desire to exert this much effort to "whack" so many "moles." And to what end?
So, carry on, friends! I've very much appreciated the general tone of the discussion. VERY much! Thank you. I wish you all the best, and I mean that. But I'm out for awhile at least.
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Jun 27, 2018 - 10:20pm PT
|
And there WILL ultimately be some human chess player that will tumble to HOW Big Blue processes, extrapolate the deterministic rules that guide it (including the ones it has 'learned') and then start systematically beating it, even given its vastly greater resources devoted to the ONE problem it is faced with.
I don't think so, and I don't believe anyone who plays chess seriously would agree with you.
|
|
|
SuperTopo on the Web
|
Let us know!