Path: ...!news.mixmin.net!weretis.net!feeder8.news.weretis.net!reader5.news.weretis.net!news.solani.org!.POSTED!not-for-mail
From: Mild Shock <janburse@fastmail.fm>
Newsgroups: comp.lang.prolog
Subject: The current Sputnik Shock (Re: NEW Logic versus OLD Logic)
Date: Mon, 6 Jan 2025 09:59:44 +0100
Message-ID: <vlg620$2964t$1@solani.org>
References: <vlat8s$26g58$3@solani.org> <vlg53q$1ssg2$1@solani.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Injection-Date: Mon, 6 Jan 2025 08:59:44 -0000 (UTC)
Injection-Info: solani.org;
	logging-data="2398365"; mail-complaints-to="abuse@news.solani.org"
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101
 Firefox/91.0 SeaMonkey/2.53.19
Cancel-Lock: sha1:0wytw1Mdp29+sA+uW9dRPWonV3s=
In-Reply-To: <vlg53q$1ssg2$1@solani.org>
X-User-ID: eJwFwQcBwEAIBDBLrGPI4Sn4l9AE6uwT5nDD4U6KupKbp1/KaVfMk14J8SexNJE72OMewy5DhS6VlvChfmNRFbc=
Bytes: 5348
Lines: 123

Hi,

The stance of LLMs towards ontology is a little
special. Asking OpenAI it tells me that LLMs can
behave as follows:

- Interpret Ontologies: Given descriptions or
   structured data from ontologies, they can
   understand and process relationships,
   classes, and properties.

- Generate Ontology Components: They can help draft
   classes, properties, or rules for ontologies when
   provided with a domain description or examples.

- Explain Ontologies: They can interpret and explain
   the structure and intent of an ontology for end-users.

I know at least of one project that tries to tap into
that, especially inputing a natural language text and
asking description components. The expected problems
are hallucination, informality and rigidity.

But since our whole education system is mainly based
on NEW Logic and not OLD Logic, which is somehow
tacitly assumed. And since LLMs have quite some grass
root dynamics, with YouTube videos platforms having

replaced the academic discussion forums and are now
influencing the terminology in the domain. The whole
matter feels like a bad joke, like waking up in a
Keeping Up with the Kardashians comedy.

Just like here:

A New Lexicon and Its Implications
https://www.computer.org/csdl/magazine/it/2024/05/10754993/21THeGJDXUI

Bye

A New Lexicon and Its Implications
https://www.computer.org/csdl/magazine/it/2024/05/10754993/21THeGJDXUI
FIGURE 5. Partial list of emergent generative AI terminology.

Mild Shock schrieb:
> Hi,
> 
> Poincare gives an interesting insight, mostly forgotten,
> that the last 100 years brought us a new logic:
> 
> Science and method
> by Poincaré, Henri, 1854-1912
> https://archive.org/details/sciencemethod00poinuoft/page/n3/mode/2up
> 
> His struggle starts at page 160, The New Logics.
> Similar Einstein was New Mechanics for him.
> 
> So what are NEW and OLD logic:
> 
> - NEW Logic: The Logic for the Past Atom Age
>    The new logic is basically a logic that allows to
>    formalize the real numbers, for example via set theory.
>    It was judged so important that the whole education
>    underwent a transformation:
> 
> One of the motives was the Sputnik shock, as a result
> of which there was a great need to catch up
> educationally in the West.
> https://de.wikipedia.org/wiki/Neue_Mathematik
> 
> - OLD Logic: The Logic for the Coming Artificial Age
>    The old logic is basically Aristotelian Begriffslogik.
>    Bis zum 19. Jahrhundert war es die dominante Tradition.
>    Die Logiker für eine solche Logik waren mehr Grammatiker
>    und Wissensimgenieure, und weniger Mathematiker die der
>    Physik zugedient haben.
> 
> https://de.wikipedia.org/wiki/Begriffslogik
> 
> LLMs give a new spin on OLD Logic, which has sent
> shockwaves into hardcore NEW Logic proponents
> as a means for Begriffslogik. Especially the use of
> of less formal means based on natural language itself,
> 
> which was somehow seen as a problem, could be a solution:
> 
> Modern ontologies:
> •    Philosophers have debated the foundations for centuries.
> •    They emphasize technical terms defined in logic.
> •    But people talk. write. and think in natural languages (NLs).
> •    Any distinctions not represented in NLs tend to be ignored.
> 
> The Great Debate between John Sowa and Barry Smith
> https://www.youtube.com/watch?v=fhYaTW5MK_U
> 
> Bye
> 
> 
> Mild Shock schrieb:
>>
>> Poincare had quite some problems with the
>> formal revolution that took place as well
>> in the last 100 or more years, starting with
>>
>> things like naive set theory and its antinomies,
>> ending with computer formalized proofs of the Keppler
>> packing nowadays. He wrote a lengthy book:
>>
>> Science and method
>> by Poincaré, Henri, 1854-1912
>> https://archive.org/details/sciencemethod00poinuoft/page/n3/mode/2up
>>
>> His struggle starts at page 160, The New Logics.
>> Similar Einstein was New Mechanics for him.
>> Mostlikely Poincaré nowadays would be a form of
>>
>> Sabine Hossenfelder with 100 YouTube videos and
>> possibly many followers. Poincaré faced the
>> destiny of any old fart that became irrelevant
>>
>> over the time and turned into a commentator.
>>
>