Re: [ontolog-forum] Watch out Watson: Here comes Amazon Machine Learning

[Thomas Johnston <tmj44p@xxxxxxx>]

Oops. "Chromodynamics", not "chronodynamics". Having recently published a book on bitemporal and tri-temporal data management, I had time on my mind.

On Friday, April 24, 2015 9:06 PM, Thomas Johnston <tmj44p@xxxxxxx> wrote:

On 4/22 (I think), John wrote:

<<<

Tom, physics is the *worst* example.  Almost nobody ever uses the
most general theories.  For any particular example, they *always*
use a special-case approximation that is tailored for that example.
And most of them, even for the same project, are *inconsistent*
with one another.

>>>

As a counterexample, I offer the standard theory of quantum chronodynamics (the unified theory of electromagnetism + the strong and weak nuclear forces). Although we have no intuitive mental image of entanglement or other "spooky" phenomena, the mathematics of quantum theory is used extensively, without any "special-case approximations" needed. It is often remarked how remarkably precise the calculations based on the standard theory are. I repeat: no approximations needed for quantum lab work.

But there is a more basic point on which John and I may or may not agree -- I'm not sure, which is why I'm continuing this exchange with John. That more basic point is this: it is the push for theoretical unification that drives science. Down at the engineering, "Let's get something built" level, we usually do not use the higher level of the currently accepted theories in our fields of interest. Which is what I take it John was pointing out.

I agree completely (with the "usually" proviso). As I said in my comment to which John's is a reply, at the "Let's get something built" level of activity, i.e. with respect to applied science, the correct procedure is indeed to use whatever tools come to hand. It is heuristics all the way, unless there is a fairly direct connection between high-level theory and engineering-level activity, as I understand there to be in the case of quantum theory/technology.

John is quite right, of course, that we do still use Newton's formulas for our work-a-day calculations, such as sending men to the moon! But Newton's theory itself was the result of a drive for theoretical unification. Newton's attitude wasn't "Well, we can calculate the trajectory of cannon balls fired at a certain angle, or the speed at which an object dropped from a certain height will hit the earth, and we can calculate the orbital path of the planets and predict eclipses and so forth, so that's good enough." Newton's theory of gravity was itself the discovery of the commonality underlying all those phenomena.

All of this seems, to me, so obvious that I suspect that John and I don't in fact have much of a disagreement. In applied science, we can and should use the conceptual tools that are at hand when those tools help us build what seem to be useful things. But the search for unification in theoretical science is not a purely "academic" exercise with no "get things done" value. New technologies, and new heuristics, arise out of new insights into how " 'things', in the widest possible sense of the term, 'hang together', in the widest possible sense of the term". (Sellars)

A non-physics example that just popped into my mind was Mendeleev's drive to organize the chemistry of his day, which resulted in the periodic table. Even without the understanding of quantum orbitals which finally explained why the periodic table has the structure it does, that table was useful. "Holes" in the table suggested new chemical elements to be discovered, and the near-by elements already in the table suggested strategies for discovering those new elements.

So if John says that he agrees with this, then the appearance of disagreement will be removed, and our conversation will not have been misleading. And if he doesn't, then I look forward to learning more about what that disagreement actually is.

On Friday, April 24, 2015 7:35 PM, Ravi Sharma <drravisharma@xxxxxxxxx> wrote:

John

Appreciate - I like description summary from you on Kuhn as to what is a good theory.

I also like the fact that sharing so far has been done through language. And language concepts have been becoming more complex, sometimes simplified by math. Meaning equivalence in multiple languages examples given are also fine.

Example: If a person did not learn formally (text) nor math, could there still be effective communication and some 'common' understanding with other person.

I am hinting to new AV capabilities, mobile and computer conversion tools and images. And we are back to Norwood Russell Hanson, Wittgenstein, etc. on how we visualize theories at least about phenomena such as planetary motions observed, etc. (of course for us on this forum, it is language, also for many it is the formula-equation about it and consequences implied, etc.)!

Rich

I could not agree more that the investments have to be recovered and for that there are models (ultimately opened like generic brands) and for those who can, open and free source are excellent. In olden times - knowledge could be acquired by only three means (Ref Sanskrit Subhashitani - roughly translated good-speech couplets):  

1. Service to Teacher
2. Money exchange
3. Barter with others (sharing of knowledge)

On the topic use of 'taxis' by you related to competition and excellence or response to stimuli, but I wanted to first understand 'how' we communicate and model agreements in our respective brains, John implies it is language, YES - as long as we agree on the concept that we are describing or experiencing!

Regards,

On Fri, Apr 24, 2015 at 7:39 AM, Rich Cooper <metasemantics@xxxxxxxxxxxxxxxxxxxxxx> wrote:

Ravi and Hans,

 

RS> I see us converging to the idea that ontology is related to "self" as individual

 

I certainly believe that is true. 

 

RS> but why could we also not have the same also shareable in some sense to be able to extend experience and learning across individuals.

Yes evolution and for that matter nature is related to differentiation, however is not there a need for synthesis to achieve common understanding?

 

What would be the cause of any statistical gradient that improves survivability of the entire population through sharing of knowledge? 

 

Anthropologists believe that we have not been competing with animals for many thousands of years, but instead that we are actively competing with other humans on a daily basis.  We compete with each other as well as help each other. 

 

Therefore, there is more of a taxis for an individual human to adapt his beliefs for matching some group for which that individual can consider his own self interest satisfiable, than there is a taxis for all people to come together as one and share without limit (music is playing ... ). 

 

Like you, I do believe it is an important goal to share knowledge with other people, but the free market already makes that happen, especially constrained by the intellectual property protections that all developed countries have created in law, and occasionally made enforceable.  That knowledge is not free; it costs money for the developer to create the knowledge, and wise national policy should encourage that such invested money should be repaid with profit in return for the development costs of the knowledge, so long as the knowledge is then shared with the population at large. 

 

The fact that knowledge can be stored in cheap computer memories means that sending the knowledge to another person is nearly free in marginal costs.  But the development costs still have to be repaid somehow or else development by the smaller more creative entities will cease.  Since they are the main source of jobs by far, that would be regrettable. 

 

Sincerely,

Rich Cooper,

Rich Cooper,

Chief Technology Officer,

MetaSemantics Corporation

MetaSemantics AT EnglishLogicKernel DOT com

( 9 4 9 ) 5 2 5-5 7 1 2

http://www.EnglishLogicKernel.com

 

From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Ravi Sharma
Sent: Friday, April 24, 2015 12:57 AM

To: [ontolog-forum]
Subject: Re: [ontolog-forum] Watch out Watson: Here comes Amazon Machine Learning - ZDNet - 2015.04.10

 

Rich and Hans

I plan to study Rich's longer message in the meantime I see us converging to the idea that ontology is related to "self" as individual but why could we also not have the same also shareable in some sense to be able to extend experience and learning across individuals.

Yes evolution and for that matter nature is related to differentiation, however is not there a need for synthesis to achieve common understanding?

Regards

 

On Thu, Apr 23, 2015 at 6:32 PM, Rich Cooper <metasemantics@xxxxxxxxxxxxxxxxxxxxxx> wrote:

Ravi and Hans,

 

Now that there are three of us, we have a group that subscribes to the belief that ontology is a mathematical description of an individual's known types of entities, plus each one's potential behaviors.  Our differences seem to be whether the ontology is generated through an individual subjective aggregation of experience, or whether the ontology is generated in some more widespread, objectively understood way. 

 

The issue seems to be how can we compute that ontology given that the individual is available for (nondestructive) reverse engineering.  If we can take what we know of ontology in the abstract, and find its wiggles, perhaps we can learn to predict the wiggles and then simulate ontologeny (!) processes versus the more well known ontogeny processes.  There may be similar underlying descriptions we can mathematize. 

 

So how can we compute an individual's ontology?  Algorithms anyone?

 

Sincerely,

Rich Cooper,

Rich Cooper,

 

Chief Technology Officer,

MetaSemantics Corporation

MetaSemantics AT EnglishLogicKernel DOT com

( 9 4 9 ) 5 2 5-5 7 1 2

http://www.EnglishLogicKernel.com

 

From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Hans Polzer
Sent: Thursday, April 23, 2015 4:16 PM

To: '[ontolog-forum] '
Subject: Re: [ontolog-forum] Watch out Watson: Here comes Amazon Machine Learning - ZDNet - 2015.04.10

 

Rich, Ravi:

 

I don’t have much to add to this dialog, but would observe that we should “embrace” diversity – and deal with that diversity, rather than suppress it. I’m not sure that encouraging growing diversity is always good – I think that’s how speciation occurs. My own belief is that we should discourage casual diversity – diversity for the sake of diversity or out of ignorance of what others have done. I will admit that sometimes that ignorance leads to positive innovation, but it also leads to “frictional losses”, to bring physics back into the discussion a bit.

 

Regarding the use of the term commonality, I’d like to point out that it requires some implicit understanding of the types of differences and the extent of such differences across which the thing in questions is “common”. However, we rarely see anyone using the term “common” to specify the differences across which something is common.

 

Hans

 

From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Rich Cooper
Sent: Thursday, April 23, 2015 5:57 PM
To: '[ontolog-forum] '
Subject: Re: [ontolog-forum] Watch out Watson: Here comes Amazon Machine Learning - ZDNet - 2015.04.10

 

Dear Ravi,

 

You wrote:

 

RS>How does math or logic help us all with different philosophic and linguistic backgrounds to help converge towards universal understanding?

 

To me, it seems that math or logic DOES NOT help us all to ... converge towards ... understanding.  Neither math nor logic helps us to understand the same things because we do not have the same experiences in our memories. 

 

RS> Can ontologies, being based on logic help us better understand neural world and mind-brain puzzle?

 

Math and logic are linear in the sense that we can use those tools to predict our near term future, or to chart the probabilities of different outcomes, and otherwise tell us something about the future given the current and the past. 

 

But people are dynamically structured, in the sense that we modify our behaviors to accomplish our individual goals.  Feedback control systems do the same thing for relatively simple mathematical structures, but again they are linear extrapolations used to squeeze the knowledge so we can find out where it leaks. 

 

The best you can say is that on some occasions, people in close groups converge to using a single word to denote a very, very simple situation.  "War" in the thirties meant WWI, then long over, not a new threat.  But in the forties, "war" supposedly still meant the same thing to everyone, but the experiences of people directly involved or affected by war were completely different than the experiences of those uninvolved and unaffected, or of Alaskan Inuit, or of Australian aborigines, or of Chinese University Professors, and many more obvious cases.  Each of these groups exist where the divergence in "understanding" is huge for critical words.  But there are many more cases that are more subtle, where the word is used by many, but the meaning is very, very different.  Hence polysemy evolved to let the divergence keep diverging. 

 

That is why I prefer to cast linguistic discussions as two agents learning how to communicate with each other, instead of casting discussions as a class of identical clones imbibing the same cool aid dictionary of a universal symbology that none of them fully understand. 

 

RS> We may have different philosophies but Newton's laws and classical mechanics go a long way for scientist and engineers; is this due to experimental verification of math that is behind them?

 

Actually, as an electrical engineer, computer scientist, and systems engineer at various points in my hectic career, I have NEVER had to use Newton's laws, classical mechanics, or any of that stuff.  I used e(t) = i(t) * z(t) and its imaginary friend in quadrature for lots of things in signal processing and electronics.  I used Fourier and Yates transforms, control theory, optimal control, directed graphs, undirected network graphs, and gazillions of other math, but I don't recall using physics or mechanics for much of anything.  I used a lot of logic in electronic logic design tasks.  I used simulation equations for many representations, but none of that is the most important math in my history. 

 

Math to model throughput, response time, multiplexing, logic, security, many other figures of merit (FOMs) get quite detailed when described in math of various kinds, but it is almost never completely linear with today's technology.  So that kind of math is far more important to my engineering associates who are not in physical mechanics, but who use math heap by heap. 

 

And, as JFS said a few emails ago, even physicists don't agree on all that physics stuff.  Look at dark matter, or cold fusion.  Every slight difference among a group of collaborating physicists creates one or more physicist with diverging opinion from the essential theories of the remaining group. 

 

So my opinion is that we necessarily have divergence in opinions for good evolutionary reasons, and we should ENHANCE that divergence at least until we understand it in full.  All progress has been made through divergent cracks in the infrastructure of that era's "common thoughts(t)". 

 

RS> A note on your exhilaration about brain being a wonderful machine, and certainly hope evolving - how did brain-mind of Mammals evolve, (we know that Aves also have some superb capabilities) and homo sapiens the best so far but universe probably (and most likely) has beings with mind-brain-consciousness billions of light years ahead of us? If we do not understand inter-species communication, how will we understand communications - among different alien life (forms)?

 

We have some small insight into a few other species, such as the work at Georgia State about primates and communications lessons learned.  There is also work (but I am not familiar with the details) of communication among wolves, dogs, crows, dolphins, parrots, whales, and many other species. 

 

But the problem, IMHO, is that each species in a conversation lacks the basic experiences and neural wiring of the other species.  Researchers call that topic "embodied intelligence". 

 

I can't hold my breath and explore the ocean rim like the dolphins do, and they can't operate mechanical and electronic equipment like I can, nor can Narwhales vote.  So our advances on that front have been really slow, but with new knowledge gradually accreting on the journal floors. 

 

Sooner or later there will emerge a pile large enough to bring us new knowledge of other species. 

 

RS>Rich - I have not given enough thought to the relationship between physics - math (logic) and philosophy hence the ambiguity. Any clarity would help.

 

For that question, I am probably not the best choice.  As I pointed out above, I have used physics very sparingly since college.  I consider math and logic to be descriptive of systems I want to understand, develop, or modify.  For that reason it is useful.  But for philosophy, I have seldom found intriguing enough philosophical questions to interest me.  You may get better responses from Matthew West, or John Sowa, or one of the others who are more interested and invested in the philosophical details. 

 

Sincerely,

Rich Cooper,

Rich Cooper,

Chief Technology Officer,

MetaSemantics Corporation

MetaSemantics AT EnglishLogicKernel DOT com

( 9 4 9 ) 5 2 5-5 7 1 2

http://www.EnglishLogicKernel.com

 

From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Ravi Sharma
Sent: Wednesday, April 22, 2015 11:55 PM
To: [ontolog-forum]
Subject: Re: [ontolog-forum] Watch out Watson: Here comes Amazon Machine Learning - ZDNet - 2015.04.10

 

Rich

 

Perhaps better choice of words for "universal understanding" could be "common" understanding or something else that implies agreement on the concept by many.

 

What I wanted to convey in the last sentence of communication was that perhaps there was a way for people to agree on understanding topics such as Physics. What is Physics? It is about description of nature of physical universe. Things about mind-brain and metaphysics are more complex or yet to be developed to the extent to which some math helps us understand physics today.

 

We may have different philosophies but Newton's laws and classical mechanics go a long way for scientist and engineers; is this due to experimental verification of math that is behind them?

 

I am looking at partitioning some parts of physics where Math can help many of us to agree on the  range of validity. Is it the mathematics (which I presume is based on Logic and or "imagination" or Concepts) that makes understanding more agreeable - often backed by measurements / experiments?

 

Rich - I have not given enough thought to the relationship between physics - math (logic) and philosophy hence the ambiguity. Any clarity would help.

 

A note on your exhilaration about brain being a wonderful machine, and certainly hope evolving - how did brain-mind of Mammals evolve, (we know that Aves also have some superb capabilities) and homo sapiens the best so far but universe probably (and most likely) has beings with mind-brain-consciousness billions of light years ahead of us? If we do not understand inter-species communication, how will we understand communications - among different alien life (forms)?

 

Can ontologies, being based on logic help us better understand neural world and mind-brain puzzle?

 

Regards,

 

On Wed, Apr 22, 2015 at 6:44 PM, Rich Cooper <metasemantics@xxxxxxxxxxxxxxxxxxxxxx> wrote:

Ravi,

 

Can you describe in much more vivid detail what you mean by your question:

 

How does math or logic help us all with different philosophic and linguistic backgrounds to help converge towards universal understanding?

 

Why do you think that "universal understanding" exists?  Perhaps there is only "individual understanding".  How would we be able to detect cases of "universal understanding" to distinguish them from "individual understanding"?

 

In my opinion, we only converge with the groups of people we interact with on a regular basis.  But we never converge for the sake of converging.  There must be an individual motivator to join a specific group.  That motivator tends to be strong in some people for some groups and the inverse for others.  But there is no good enough history database to really collect actual data on the hundreds of years of written history behind us.  But experiments like that could be done more easily now. 

 

So we can only sigh at the lack of history details.

 

Sincerely,

Rich Cooper,

Rich Cooper,

 

Chief Technology Officer,

MetaSemantics Corporation

MetaSemantics AT EnglishLogicKernel DOT com

( 9 4 9 ) 5 2 5-5 7 1 2

http://www.EnglishLogicKernel.com

 

From: ontolog-forum-bounces@xxxxxxxxxxxxxxxx [mailto:ontolog-forum-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Ravi Sharma
Sent: Wednesday, April 22, 2015 2:22 PM
To: [ontolog-forum]
Subject: Re: [ontolog-forum] Watch out Watson: Here comes Amazon Machine Learning - ZDNet - 2015.04.10

 

John

YES -

1. All our recent or useful discoveries and products are based on engineering which are ultimately dependent on physics approximations. Physics attempts to include life sciences phenomena but not yet convincingly till we can synthesize Life reliably! But otherwise physics is attempt at description of at least non-life matter.
2. Scientists / Engineers know their models are based on Range of Validity and approximations related to desired accuracy, deviations from mean, etc.
3. Reality and truth get us into the fuzzy areas where knowledge of how the brain works could help us better define the context or meaning. All Cosmic skylight (e.g. at night) falling on retina - does it describe reality? What kind?

• when individual photons from different sources impinged on retina but actually originated from different objects at different times some of which in our local-time may not even exist now.
• how long after photon entered retina - i.e. to individual subject's brain processing time?
• as believed in some philosophies that what appears in senses is not-real the reality is Only One.Thomas Johnson's description in email thread: that Being is One (and so an explanation of Being should be one) it goes earlier to Parmenidian - centuries earlier than 600BC.
• your earlier comments relating to models of objects perceived by individual brain and connectionism and including referenced URL- your work with Majumdar.

4. Now my Question - How does math or logic help us all with different philosophic and linguistic backgrounds to help converge towards universal understanding? At least Physicists understand Relativity and Symmetry models through mathematical "language"?

Regards,

Ravi

 

On Wed, Apr 22, 2015 at 12:46 PM, John F Sowa <sowa@xxxxxxxxxxx> wrote:

On 4/22/2015 2:45 PM, Thomas Johnston wrote:
> But two theories are not better than one, as regimented attempts
> to understand things. I think the underlying intuition which pushes
> physicists towards a unified theory...

Tom, physics is the *worst* example.  Almost nobody ever uses the
most general theories.  For any particular example, they *always*
use a special-case approximation that is tailored for that example.
And most of them, even for the same project, are *inconsistent*
with one another.

Physicists have known for over a century that Newtonian physics
is only an approximation, but it is still the most widely used
theory.  But even then, there are huge numbers of special cases
of Newtonian mechanics:  supersonic fluids; subsonic fluids;
turbulent flow; viscous flow; incompressible fluids (which really
aren't).  The biggest examples are the incredible number of
approximations for computing the global weather -- different
versions for multiple levels of the atmosphere, different regions
of the earth, different terrains, geographies, ocean currents,
times of day, seasons of the year, etc., etc., etc...

The total number of widely used approximations is in the thousands.
The number of detailed approximations is in the billions -- every
engineer for every project takes a large number of general-purpose
approximations and specializes them for different parts of the project.

Every large system -- ranging from your cell phone to your car to the
trains, planes, and road systems you use every day -- is based on a
large collection of mutually inconsistent approximations to the basic
laws of physics -- all of which are *known* to be false when pushed
to the limits.

Fundamental principle:  The human brain is the most complex natural
system known.  It is far more complex than the global weather, the
Large Hadron Collider, or the global collection of all the human
constructions on earth.

Analogy:  The Greek theories of the cosmos by the pre-Socratics
are closer to modern physics than any current theory of the brain
is to the way it actually works.

John

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J

--

Thanks.
Ravi
(Dr. Ravi Sharma)
313 204 1740 Mobile

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
 

--

Thanks.
Ravi
(Dr. Ravi Sharma)
313 204 1740 Mobile

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
 

--

Thanks.
Ravi
(Dr. Ravi Sharma)
313 204 1740 Mobile

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
 

--

Thanks.
Ravi
(Dr. Ravi Sharma)
313 204 1740 Mobile

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (01)