Hi,
About Deep learning. Do we really need
GPU and Torch libraries, aren't there
other methods do arrive at the same.
Do we really need to invest in nuclear
power plant and shove money up a
California based company as this
"historical" photo suggests (could be fake):
Some pics from when Jensen delivered the
first @Nvidia AI system to @OpenAI
https://twitter.com/elonmusk/status/1759295781196927438
What is the key to modern Knowledge
Management, how does modern Knowledge
Engineering work. Wasn't there a perfection
of outsourcing and text annotation jobs.
Somehow this can be accelerated by subtle
social media methods, that allow a form
of hidden crowd sourcing? And like 80%
of the civil non military scientific community
has absolutely no clue whats going on?
Bye
Mild Shock schrieb:
Hi,
But I don't have a final solution. I am
still a seeker, a free seeker free from
legacy and orthodoxy, using free speech.
Just like Luce:
The official mascot for the Catholic
Church’s 2025 Jubilee Year is named
“Luce,” which is Italian for “light.” https://www.catholicnewsagency.com/news/260129/meet-luce-the-vatican-s-cartoon-mascot-for-jubilee-2025
LoL
Bye
Mild Shock schrieb:
Hi,
Whats also extremly cringe is the
struggel that Scryer Prolog is experiencing.
It has adopted a bloat of post Herbrand
Prolog nonsense, due to Colmerauer, namely
fucking cyclic term unification. Its
crashing under its own bloath of nonsense,
just go to GitHub and count the issues.
Last time I looked it had like 200 issues,
and now after a blink its already at 300 issues.
335 Open - Scryer Prolog - Issues
https://github.com/mthom/scryer-prolog/issues
WTF?
Bye
Mild Shock schrieb:
SICStus Prolog is overrated. A lot of things
in Prolog logic programming are overrated, so
that it has lost it to a lot of turf over
the last decade. But of course clinging to
an old code base, and reselling the same old
shit over and over again, is a very lucrative
business compared to real innovation.
Until OpenAI comes and does LLM. I am pretty
sure we will see some blocks stumbling. For
example the paper I was citing was traditional
machine lerning, it was even not deep learning.
Mostlikely Prolog will be hit much more than
it is already hit with Python & Co. taking over
a lot of data science turf.
LoL
Julio Di Egidio schrieb:
On 06/11/2024 09:51, Mild Shock wrote:
What the fuck does this mean:
The Nazi-retarded shithole is all but going to fix itself.
Why is change_arg/3 not common among Prolog systems.
Only poisoned meatballs for the public.
https://github.com/SWI-Prolog/swipl-devel/issues/1328
-Julio
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)