Alex,
Absolutely NOT!!!!!
Alex> we should keep diagrams separately
Diagrams are the intermediate step between continuous images and linear notations. The
first step in analyzing any kind of continuous space or structure is to draw a diagram.
Just look at any book on geometry, starting with Euclid. (There were earlier manuscripts.
But after Euclid, nobody bothered to recopy them.)
Just look at any map. The first step between a continuous landscape and any notation of
any kind is a map that identifies cities, towns, mountains, rivers, streets, and many,
many POINTS and LINES of interest,
And look at any plan for any kind of structure -- airplanes, buildings, cars, bridges,
etc., etc. etc. Every continuous design begins with a map or blueprint or other kind of
diagram.
A diagram is a geometrical structure made of discrete points and lines that map to
significant points and lines on a continuous space. That is the fundamental reason for
diagrams in knowledge representation.
The most ancient monuments like Stonehenge and other structures on all continents except
Antarctica are diagrams that reflect the significant points (stars, planets, and the moon)
in a continuous sky. The constellations that people imagine are based on diagrams with
lines that connect those points.
That is why I keep repeating the importance of diagrams. People who are born deaf can
communicate perfectly well with moving three-dimensional diagrams. Children who have
normal hearing, but are born to deaf parents babble with the hands, not by sounds. Their
first languages are sign languages -- and they are not handicapped in any way when they
finally learn to talk.
There is strong evidence that human spoken languages evolved from the sign languages of
the apes. In fact, hearing people who learn signed languages score higher on IQ tests
that involve diagrams -- after they learn a sign language.
That is one of many reasons why I keep emphasizing: LLMa are NOT a step toward a
human-level intelligence. They're a useful adjunct, but not a fundamental form of
knowledge representation. Diagrams are fundamental, and linear notations are useful for
(a) typing, and (b) saving space on a printed page..
John
PS: There were some very smart, but hopelessly misguided mathematicians called the
Bourbaki. They tried to get rid of diagrams. They wrote some interesting books, but
their goal of getting rid of diagrams was hopelessly misguided. And it failed miserably.
Their books still contain some useful ideas, but nobody uses them tp teach students..
----------------------------------------
From: "Alex Shkotin" <alex.shkotin(a)gmail.com>
Ravi,
Relationships between World of Words and World of Pictures (3D by the way) is a great
topic. We will need a new thread to collect ideas.
But we should keep diagrams separately, IMHO :-)
Alex
вс, 12 нояб. 2023 г. в 12:06, Ravi Sharma <drravisharma(a)gmail.com>om>:
Alex
Thiswas a 1-2 page written exchange with the then living philosopher Professor Norwood
Hensen ay Yale. He was also known as a flying Professor and unfortunately he died young
while flying.
The summary of my assertion with him was that what we model in our mind and understand as
the meaning is often based on a picture or visual construct and not always the text to
understand. We can talk about this later, it is also studied in infants or children who
inherently understand many things such as fall due to gravity, etc,
Now as we are working in ontology, we can relate these days with graphs, knowledge graphs
and AI beyond LLM!
Thanks for showing interest in my old theory, I will try to see in my archives if I based
on what he then published anywhere that I read and responded or how we corresponded with
each other, After 3 years at Florida I went to Yale for postdoctoral research and also did
teaching in the physics Department.
Thanks.Ravi