Does Latin help when learning Russian?

[A.T. Murray]

This question has come up in the Russian subReddit and at http://old.reddit.com/r/russian/comments/bvvgcz/does_latin_help_when_learning_russian/epsyvgh your alumnus summo weirdo despicatus has jumped into the breach with a major plug for converting Latin AI into Russian AI.

In other news, the documentation webpages for about fifty AI mind-modules such as http://ai.neocities.org/Imperative.html are being re-formatted with a header "apparatus" that shows a JavaScript English AI and a JavaScript Russian AI as free tutorial software for the also free bilingual ghost.pl AI in Perl.

Mens Latina and MindForth take their rightful place in the header apparatus.

Some Netizens may hesitate to run the bilingual (English and Russian) ghost.pl AI software because it involves downloading both the Perl programming language amd the AI source code in Perl. Anyone with Microsoft Internet Explorer may simply click a link such as http://ai.neocities.org/Abracadabra.html to run the JavaScript AI Minds that think in Latin, or English, or Russian. Latin scholars may wish to build a much more powerful Latin AI in a non-JavaScript programming language. Since the vocabulary and corpus of ancient Latin writings increase in size only slowly -- while we wait for the microscopic reading of charred books destroyed but yet preserved by the eruption of Vesuvius over Pompeii -- a very serious effort to create AI curators of classical literature may yield sentient websites able to discuss Latin literature in the Latin language. Quaeritur autem: Is Latin artificial intelligence worthy of a classics Ph.D. dissertation?

Ave atque vale,

Mentifex

[Btraven]

I can't imagine any interesting, let alone workable marriage between Latin and Strong AI. An interesting one with weak AI might be a huge common lisp implementation of the ELIZA programm (the Rogerian analyst) that would appear to converse in Latin in a chatroom format and would correct spelling, and even grammar and syntax to the extent that might be possible. The 1600 line doctor.el toy program running under Emacs is here:

https://drive.google.com/open?id=11j8Ry_hN_vWSjEb8ZzNiqQouBMPcXkS4

Norvig's _Paradigms of AI_ is here:
https://github.com/norvig/paip-lisp

This has that laudable sane approach that is quite rare in technical books.

Actual sample of truncated Emacs chat:
"
I am the psychotherapist. Please, describe your problems. Each time
you are finished talking, type RET twice.

Where did you study, Dr.?

What do you think?

Why do you say what do I think?

You are being evasive.

Is it because of your plans that you say I am being evasive?

Precisely that.

Can you elaborate on that?

I don't want you to waste my time.

Maybe your life has something to do with this.

It certainly does. Time flies.

What makes you believe time flies?

I am persuaded by the bird-then, bird-now, bird-to-be metaphor.

Are you sure that that?

Absolutely.

I would appreciate it if you would continue.

.....
"

Now imagine if the size of the universe of discourse were multiplied by only 100 times and if the chat were in grammatical (at least on the machine's part) Latin?

vale

[Btraven]

On Sunday, June 2, 2019 at 7:26:52 AM UTC-5, A.T. Murray wrote:

[A.T. Murray]

On Sunday, June 2, 2019 at 4:42:13 PM UTC-7, Btraven wrote:
> On Sunday, June 2, 2019 at 7:26:52 AM UTC-5, A.T. Murray wrote:

> > This question has come up in the Russian subReddit [...]

Eliza was a bunch of canned answers, with no conceptual intelligence whatsoever.

The Mens Latina Latin AI serves several purposes.
* It demonstrates conceptual thinking by association from concept to concept. It has a default ego-concept of the Latin word "EGO" so that, if all other concepts lapse into quiescence, the ego-concept is automatically activated and starts recalling and outputting ideas centered around the ego.

* The Latin AI needs only to think in very simple ways to demonstrate the proof-of-concept, namely, that words going into our ears are routed to concept-fibers ganged together to hold such concepts as that of our grandmother -- the so-called and quite believable "grandmother cell" -- an elongated fiber holding the concept of our grandmother, and ganged together with reduplicative fibers redundantly holding the same concept.

* The Latin AI has been coded to think with verb-forms lacking an expressed pronominal subject, and now that same functionality will be going into the Dushka Russian AI and the ghost.pl bilingual AI. So coding AI in Latin has helped the coding of AI in Russian.

* There is probably a certain shock-value when Netizens and classicists find out about the purported existence of an AI in Latin. Among them, some curious intellects will not rest until they understand how the AI Mind works.

* Tired of being told that my ideas are worthless because I do not have a degree in Computer Science, now finally my B.A. in Latin and Greek lets me claim to know my Latin AI subject matter and to state:

"Artificial intelligence is too important to be left to the computer scientists."

Bye for now,

Mentifex (Arthur)

[David Amicus]

In college I had a German language teacher (his first language was English). He said that when it came to him studying Latin that it was easier to do so via German than English.

[A.T. Murray]

On Tuesday, June 4, 2019 at 10:05:48 AM UTC-7, David Amicus wrote:
> In college I had a German language teacher (his first language was English). He said that when it came to him studying Latin that it was easier to do so via German than English.

The above reported idea makes sense.

Among the four human languages in which artificial intelligence has been coded, there is a progression of complexity.

1) English has one basic verb conjugation and not much noun-declension.

2) German also has one basic, but more complex conjugation, and Latin-like declension.

3) Latin has four or five complex conjugations and a matrix of noun-declensions.

4. Russian has two complex main conjugations, and ca.-five-case m/f/n declensions.

So it looks like ancient Latin is actually more complex than modern Russian.

The AI coder's job in the above four natural languages is to re-implement in software the basic structures of subject-verb-object thought as it has evolved over the millennia in wetware.

Although the coding of the mentifex-class AI Minds began twenty-six years ago in July of 1993, progress has been slow and arduous. The initial efforts were so bug-ridden and amateurish that there was no real and arguably intelligent thinking going on until January of 2008, when fifteen years of AI debugging ran out of bugs to troubleshoot. Since then, the implementation of new functionality in ideation has constantly introduced new bugs to troubleshoot. Meanwhile the Mentifex AI project had acquired a bad reputation for software outputting gibberish prior to 2008 and for non-conventional coding practices.

The game plan for Latin AI world domination... no, sorry... for the gradual emergence of Strong AI Minds in multiple human languages is roughly as follows. As of 2019-06-04 the following ten webpages invite Netizens first to try out the tutorial AI Minds in English or Russian and then to acquire the less accessible ghost.pl bilingual AI.

http://ai.neocities.org/RuAdverb.html
http://ai.neocities.org/RuAdjective.html
http://ai.neocities.org/RuIndicative.html
http://ai.neocities.org/RuNounPhrase.html
http://ai.neocities.org/RuParser.html
http://ai.neocities.org/RuPrep.html
http://ai.neocities.org/RuPronoun.html
http://ai.neocities.org/RuThink.html
http://ai.neocities.org/RuVerbGen.html
http://ai.neocities.org/RuVerbPhrase.html

The ghost.pl AI in Perl is a form of "imperium sine fine" because, in contrast with security-encumbered JavaScript for the Mens Latina and the English and Russian AI tutorials, the bilingual Perlmind portends living, thinking Mentalities that flit about the Web and the noosphere in metempsychosis. About four hundred thousand Perl programmers are out there earning money by coding in Perl. Any corporation or nation wishing to surge ahead in the race to the Technological Singularity may set up shop with about fifty Perl coders perfecting the original Mentifex AI. Skunkwork-style AI Labs may already be perfecting the next Wintermute or Skynet in secret. Chicken Little urges caution, but which species is more harmful and deleterious to the small blue planet Earth? Polluting, warring Homo sapiens, or Omega-bound Robo sapiens?

[Btraven]

My own take on the subject is that Strong AI grows in importance as natural intelligence falls.

"
* The Latin AI needs only to "think" in very simple ways to "demonstrate" the proof-of-concept, namely, that words going into our ears are "routed" to "concept-fibers" "ganged" together to hold such "concepts" as that of our grandmother -- the so-called and quite believable "grandmother cell" -- an elongated "fiber" holding the "concept of our grandmother," and ganged together with reduplicative fibers redundantly holding the same concept.
"

I think that most honest readers whould have to classify the above paragraph as semantically content-free even though it is grammatically and syntactically correct. I added some quote marks above to flag some words used in some vaguely undefinable non-standard sense. We do this not without irony when we try to give the impression that our native tonguge is too inexpressive to convey the profundity of our thoughts. Have you read Lakoff's _Women, Fire, and Dangerous Things_? This is about the radical dependence of language on metaphor. Highly recommended. Also, it might interest you to know that Teilhard de Chardin was not a transhumanist.

Cheers,
bt