Re: Is this program OOP?

On Thu, 12 Dec 2024 09:59:14 +0200
Paavo Helde <eesnimi@osa.pri.ee> wibbled:

On 11.12.2024 10:40, Muttley@DastardlyHQ.org wrote:

We had to use it in the AI module at uni. :/
 
I suspect its raison d'etre has rather been shot down these days as AI has
moved away from propositional logic to neural networks.

>
Some propositional logic would be very beneficial for the current
babble-GPT style AI. This is exactly what they are missing. Alas, this
would assume they understand a bit what they are babbling about, which
is not the case.

The problem with propositional logic is its completely restricted to what
its been programmed with. If there was a way for a logic system to slurp the
internet the way neural nets do and create its own clean logical deductions it
might work but clearly there isn't or someone would have done it by now.

Given that even many people have huge problems with a), b) and d) I
would not hold my hopes high that all these things would somehow
randomly emerge.

Given that even the programmers don't really understand how these systems
work at the deep level I don't think anyone can make any predictions about
wghat they might do. However the limit of the current designs seems to have
been reached beyond tweaks here and there.