חדש באתר: עוזר בינה מלאכותית המבוסס על כתביו ושיעוריו של הרב מיכאל אברהם

Q&A: AGI as a Thinker

Back to list  |  🌐 עברית  |  ℹ About
Originally published:
This is an English translation (via GPT-5.4). Read the original Hebrew version.

AGI as a Thinker

Question

A new update to ChatGPT has been released.
The innovation: it is capable of solving mathematical equations (beyond previous models, which were poor at this. Broadly speaking, because they relied on information that had been fed into them, without a process of reasoning).
In other words, it performs better than a mathematics student on exams,
and for the first time it can pass the matriculation exam successfully.a0
(It was tested!)
 
The developer was interviewed and said that we are two years away from AGI—artificial general intelligence,
which will be better than humans in *every* field.
 
Meaning, on every test, in every possible discipline, it will get a higher and more perfect score than any human being.a0
 
In mathematics, physics, and many other fields it is already better.
 
Do you think it is possible that it will be able to produce ideas, philosophy, and theology that are finer / more precise / more fascinating / more comprehensive than those of human beings?
 
If it can solve equations more complicated than Euler and Schulz,a0
will it be able to produce ideas superior to Kant’s?
 
The mechanism by which great human beings produce ideas and philosophy,

is it replaceable by software that is smarter than human beings in every field?

Answer

I do not know. I have devoted columns to describing what I think may be the limitations of such a golem. Technical abilities are almost certainly only a matter of time. The big question is whether all of our abilities are technical. See columns 35, 590591.

Leave a Reply

Back to top button