From Newsgroup: comp.lang.prolog

Hi,

The cest pool of bit rot is finally gone:

(Bit rot, because they did everything in a
facist way ignoring your copyright to prevent
you from deleting you own out dated stuff)

CoincidenceIDontThinkSo
https://9gag.com/gag/abAG4oE

The Fall of Stack Overflow https://observablehq.com/@ayhanfuat/the-fall-of-stack-overflow

The downfall is related to the appearance of
ChatGPT. I feel pitty for Ulrich Neumerkel false
and others who spammed stackoverflow for years

with their infinite dif/2 nonsense.

Bye

Mild Shock schrieb:

Hey folks, please take note:

"Rust is the last programming language made for humans
that we'll need. Future languages will be optimized for
machines, where AI handles all the coding" https://twitter.com/nath_simard/status/1816147809608851830

Sounds like utter bullshit to me. Must be a
stupid AI, that cannot do the the safety precautions
of Rust. Why would I need rust, if I have an AI?

For this exampe, the AI might generate:

#include <stdio.h>

int main(int argc, char* argv[]) {
   printf("Hello %s\n", argv[0]);
}

Which might indeed generate a segfault. In
case there are zero arguments, argv[0] might
hafve the value NULL. https://www.tu-chemnitz.de/urz/archiv/kursunterlagen/C/kap3/argmain.htm

So why blame the target language, and not the AI?
This is a red herring argument in favor of Rust.

--- Synchronet 3.20a-Linux NewsLink 1.114

From Newsgroup: comp.lang.prolog

But ChatGPT has still to learn a lot, it
cannot make the link from Prolog to
lambda-Prolog to natural deduction.

Somehow after some questioning ChatGPT
presented me this Prolog code:

likes(john, pizza).
likes(mary, sushi).
likes(sam, burgers).

friend(john, mary).
friend(mary, sam).

likes_what_friends_like(Person, Thing) :-
friend(Person, Friend),
likes(Friend, Thing).

I then asked for a Prolog query that
solves the following question:

"who should john befriend so that we
can assume he likes burgers"

Interestingly ChatGPT strated using
Prolog and natural deduction on the meta
level, mentioning modus ponense, and answering

Proof via Natural Deduction:
Assumption: Assume friend(john, X) (where X = sam).

Apply the Rule: Since friend(john, sam) is true, and likes(sam, burgers)
is also true, we apply the rule likes_what_friends_like(john, Thing) and
infer likes_what_friends_like(john, burgers).

Conclusion: Therefore, likes_what_friends_like(john, burgers) holds,
which implies that John would like burgers if he befriends Sam.


Mild Shock schrieb:

Hi,

The cest pool of bit rot is finally gone:

(Bit rot, because they did everything in a
facist way ignoring your copyright to prevent
you from deleting you own out dated stuff)

CoincidenceIDontThinkSo
https://9gag.com/gag/abAG4oE

The Fall of Stack Overflow https://observablehq.com/@ayhanfuat/the-fall-of-stack-overflow

The downfall is related to the appearance of
ChatGPT. I feel pitty for Ulrich Neumerkel false
and others who spammed stackoverflow for years

with their infinite dif/2 nonsense.

Bye

--- Synchronet 3.20a-Linux NewsLink 1.114

From Newsgroup: comp.lang.prolog

My objection was then that Prolog didn't do
the proof only ChatGPT did it.

"But Prolog is supposed to search for proofs.
What is the proof?"

It then continued with attempts like:

what_if_befriend(Person, Friend, Goal) :-
assert(friend(Person, Friend)), % Assert the new friendship
Goal, % Check the goal with the new friendship
retract(friend(Person, Friend)). % Retract after checking the goal

Only the above will not work, since
assert/1 cannot return a variable binding.

So what is the magic behind lambda-Prolog
and hypothetical reasoning? lambda-Prolog
not being in the scope of ChatGPT?

Mild Shock schrieb:

But ChatGPT has still to learn a lot, it
cannot make the link from Prolog to
lambda-Prolog to natural deduction.

Somehow after some questioning ChatGPT
presented me this Prolog code:

likes(john, pizza).
likes(mary, sushi).
likes(sam, burgers).

friend(john, mary).
friend(mary, sam).

likes_what_friends_like(Person, Thing) :-
    friend(Person, Friend),
    likes(Friend, Thing).

I then asked for a Prolog query that
solves the following question:

"who should john befriend so that we
can assume he likes burgers"

Interestingly ChatGPT strated using
Prolog and natural deduction on the meta
level, mentioning modus ponense, and answering

Proof via Natural Deduction:
Assumption: Assume friend(john, X) (where X = sam).

Apply the Rule: Since friend(john, sam) is true, and likes(sam, burgers)
is also true, we apply the rule likes_what_friends_like(john, Thing) and infer likes_what_friends_like(john, burgers).

Conclusion: Therefore, likes_what_friends_like(john, burgers) holds,
which implies that John would like burgers if he befriends Sam.

Mild Shock schrieb:

Hi,

The cest pool of bit rot is finally gone:

(Bit rot, because they did everything in a
facist way ignoring your copyright to prevent
you from deleting you own out dated stuff)

CoincidenceIDontThinkSo
https://9gag.com/gag/abAG4oE

The Fall of Stack Overflow
https://observablehq.com/@ayhanfuat/the-fall-of-stack-overflow

The downfall is related to the appearance of
ChatGPT. I feel pitty for Ulrich Neumerkel false
and others who spammed stackoverflow for years

with their infinite dif/2 nonsense.

Bye

--- Synchronet 3.20a-Linux NewsLink 1.114

From Newsgroup: comp.lang.prolog

How it started:

An exciting day! @spolsky called me and let me
know. Today's sale of Stack Overflow, most importantly,
lets Stack Overflow continue as an independent site --
and also mints 61 new millionaires.
Jeff Atwood - 6:47 PM · Jun 2, 2021 https://twitter.com/codinghorror/status/1400132013068218370

How its going:

Stack Overflow cuts 28% of its staff https://techcrunch.com/2023/10/17/stack-overflow-cuts-28-of-its-staff/

Mild Shock schrieb:

My objection was then that Prolog didn't do
the proof only ChatGPT did it.

"But Prolog is supposed to search for proofs.
What is the proof?"

It then continued with attempts like:

what_if_befriend(Person, Friend, Goal) :-
    assert(friend(Person, Friend)),  % Assert the new friendship
    Goal,  % Check the goal with the new friendship
    retract(friend(Person, Friend)).  % Retract after checking the goal

Only the above will not work, since
assert/1 cannot return a variable binding.

So what is the magic behind lambda-Prolog
and hypothetical reasoning? lambda-Prolog
not being in the scope of ChatGPT?

--- Synchronet 3.20a-Linux NewsLink 1.114

From Newsgroup: comp.lang.prolog

Hi,

Yes exactly, they shoot themselves into the foot.
They adopted the academic closed world attitute,
as can be seen here, not noticing that the John

Doe programmers are not necessarily a theoretical
computer science titanic A-class member:

As an example, consider this question – invented, but not far-fetched: "Random keeps giving me the same numbers. Is it broken?"
This is a low-quality question, in my view. https://codeblog.jonskeet.uk/2018/03/17/stack-overflow-culture/

Now try the same with ChatGPT:

Q: Random keeps giving me the same numbers. Is it broken?

A: It sounds like your Random object might not be
re-seeded, which could cause it to generate the same
sequence of numbers every time you run your program.
By default, Random is seeded with the current time in
milliseconds, but if you create it multiple times in
quick succession, it might receive the same seed and
generate the same numbers.

Bye

Mild Shock schrieb:

Hi,

The cest pool of bit rot is finally gone:

(Bit rot, because they did everything in a
facist way ignoring your copyright to prevent
you from deleting you own out dated stuff)

CoincidenceIDontThinkSo
https://9gag.com/gag/abAG4oE

The Fall of Stack Overflow https://observablehq.com/@ayhanfuat/the-fall-of-stack-overflow

The downfall is related to the appearance of
ChatGPT. I feel pitty for Ulrich Neumerkel false
and others who spammed stackoverflow for years

with their infinite dif/2 nonsense.

Bye

--- Synchronet 3.20a-Linux NewsLink 1.114

From Newsgroup: comp.lang.prolog

Hi,

I personally like that ChatGPT tolerates all my
sloppiness, since I am always in a hurry, don't
have time for some bullshit.

And I pay the bill by myself, if ChatGPT doesn't
understand me. I don't need some moderator to
force some ideal into my throat, anyway

most of the time ChatGPT understands me
nevertheless. Has probably to do how
the artificial intelligenc of ChatGPT works:

Attention Is All You Need
https://arxiv.org/abs/1706.03762

But stack overflow gets also drained by other
developments like for example:

- Thousends of tutorial web sites:
There is a real cottage industry, of
commercial web sites, that explain you
web programming etc.. They are commercial
since they show banner ads.

- Thousends of tutorial videos:
There is a similar cottage industry
for videos. Also commercial because
the show ads. Even google search
doesn't show stack overflow as first
result anymore.

- Declining Business Model Job Search
With "low code", the number of programmer
jobs as we know it will go down. A website
specialized on job advertising will automatically
loose revenue:

https://stackoverflow.co/advertising/

Bye

Mild Shock schrieb:

Hi,

Yes exactly, they shoot themselves into the foot.
They adopted the academic closed world attitute,
as can be seen here, not noticing that the John

Doe programmers are not necessarily a theoretical
computer science titanic A-class member:

As an example, consider this question – invented, but not far-fetched: "Random keeps giving me the same numbers. Is it broken?"
This is a low-quality question, in my view. https://codeblog.jonskeet.uk/2018/03/17/stack-overflow-culture/

Now try the same with ChatGPT:

Q: Random keeps giving me the same numbers. Is it broken?

A: It sounds like your Random object might not be
re-seeded, which could cause it to generate the same
sequence of numbers every time you run your program.
By default, Random is seeded with the current time in
 milliseconds, but if you create it multiple times in
quick succession, it might receive the same seed and
generate the same numbers.

Bye

Mild Shock schrieb:

Hi,

The cest pool of bit rot is finally gone:

(Bit rot, because they did everything in a
facist way ignoring your copyright to prevent
you from deleting you own out dated stuff)

CoincidenceIDontThinkSo
https://9gag.com/gag/abAG4oE

The Fall of Stack Overflow
https://observablehq.com/@ayhanfuat/the-fall-of-stack-overflow

The downfall is related to the appearance of
ChatGPT. I feel pitty for Ulrich Neumerkel false
and others who spammed stackoverflow for years

with their infinite dif/2 nonsense.

Bye

--- Synchronet 3.20a-Linux NewsLink 1.114

From Newsgroup: comp.lang.prolog

Hi,

Also Stack Overflow is very bad in reading
the signs of the times:

Generative AI is now integral to the
technologist workflow, yet many developers
are still skeptical. According to our 2024
#Developer Survey, 79% of respondents cited
misinformation as their primary ethical
concern regarding AI.
https://x.com/stackoverflow/

The survey is not so bad:
https://survey.stackoverflow.co/2024/ai

So they have it right in front of their
nose that they should shut down

their grammar nazi nonsense and find
new ways, alternative ways to harness

and curate knowlegde?

Bye

Mild Shock schrieb:

Hi,

I personally like that ChatGPT tolerates all my
sloppiness, since I am always in a hurry, don't
have time for some bullshit.

And I pay the bill by myself, if ChatGPT doesn't
understand me. I don't need some moderator to
force some ideal into my throat, anyway

most of the time ChatGPT understands me
nevertheless. Has probably to do how
the artificial intelligenc of ChatGPT works:

Attention Is All You Need
https://arxiv.org/abs/1706.03762

But stack overflow gets also drained by other
developments like for example:

- Thousends of tutorial web sites:
  There is a real cottage industry, of
  commercial web sites, that explain you
  web programming etc.. They are commercial
  since they show banner ads.

- Thousends of tutorial videos:
  There is a similar cottage industry
  for videos. Also commercial because
  the show ads. Even google search
  doesn't show stack overflow as first
  result anymore.

- Declining Business Model Job Search
  With "low code", the number of programmer
  jobs as we know it will go down. A website
  specialized on job advertising will automatically
  loose revenue:

https://stackoverflow.co/advertising/

Bye

Mild Shock schrieb:

Hi,

Yes exactly, they shoot themselves into the foot.
They adopted the academic closed world attitute,
as can be seen here, not noticing that the John

Doe programmers are not necessarily a theoretical
computer science titanic A-class member:

As an example, consider this question – invented, but not far-fetched:
"Random keeps giving me the same numbers. Is it broken?"
This is a low-quality question, in my view.
https://codeblog.jonskeet.uk/2018/03/17/stack-overflow-culture/

Now try the same with ChatGPT:

Q: Random keeps giving me the same numbers. Is it broken?

A: It sounds like your Random object might not be
re-seeded, which could cause it to generate the same
sequence of numbers every time you run your program.
By default, Random is seeded with the current time in
  milliseconds, but if you create it multiple times in
quick succession, it might receive the same seed and
generate the same numbers.

Bye

Mild Shock schrieb:

Hi,

The cest pool of bit rot is finally gone:

(Bit rot, because they did everything in a
facist way ignoring your copyright to prevent
you from deleting you own out dated stuff)

CoincidenceIDontThinkSo
https://9gag.com/gag/abAG4oE

The Fall of Stack Overflow
https://observablehq.com/@ayhanfuat/the-fall-of-stack-overflow

The downfall is related to the appearance of
ChatGPT. I feel pitty for Ulrich Neumerkel false
and others who spammed stackoverflow for years

with their infinite dif/2 nonsense.

Bye

--- Synchronet 3.20a-Linux NewsLink 1.114

From Newsgroup: comp.lang.prolog

Or just give up Q&A sites all together. They
live in a bubble that they sit on a treasure.

Any clue why? Simply because text will pretty soon
not be mandatory anymore for training of Chat bots:

Early 2000s: Automatic transcription services were
rudimentary, relying on basic speech recognition
software with limited accuracy.

2010s: The technology began to improve significantly
with the development of deep learning and neural
networks. Platforms like YouTube introduced automatic
captions in 2006, but they were often unreliable
and required manual corrections.

2016–2018: Major tech companies, including Google and
Microsoft, launched more advanced speech-to-text
services. These platforms utilized machine learning
algorithms and large datasets to provide more
accurate transcriptions.

2020–2023: With advancements in AI and natural language
processing (NLP), transcription services became much
more accurate and widely available. Companies like Rev,
Otter.ai, and others began offering services with high
accuracy and specialized features (such as speaker
identification, real-time transcription, and more).

But ChatGPT isn't there yet?

Although ChatGPT itself is not directly trained
on video data, multimodal models (such as GPT-4's
ability to process images) are paving the way for
broader applications of video-based learning in AI.

So the cottage industry of training videos
will be the next target of knowledge harvesting?
Directly incorporating your professors lectures?

LoL

Mild Shock schrieb:

Hi,

Also Stack Overflow is very bad in reading
the signs of the times:

Generative AI is now integral to the
technologist workflow, yet many developers
are still skeptical. According to our 2024
#Developer Survey, 79% of respondents cited
misinformation as their primary ethical
concern regarding AI.
https://x.com/stackoverflow/

The survey is not so bad:
https://survey.stackoverflow.co/2024/ai

So they have it right in front of their
nose that they should shut down

their grammar nazi nonsense and find
new ways, alternative ways to harness

and curate knowlegde?

Bye

Mild Shock schrieb:

Hi,

I personally like that ChatGPT tolerates all my
sloppiness, since I am always in a hurry, don't
have time for some bullshit.

And I pay the bill by myself, if ChatGPT doesn't
understand me. I don't need some moderator to
force some ideal into my throat, anyway

most of the time ChatGPT understands me
nevertheless. Has probably to do how
the artificial intelligenc of ChatGPT works:

Attention Is All You Need
https://arxiv.org/abs/1706.03762

But stack overflow gets also drained by other
developments like for example:

- Thousends of tutorial web sites:
   There is a real cottage industry, of
   commercial web sites, that explain you
   web programming etc.. They are commercial
   since they show banner ads.

- Thousends of tutorial videos:
   There is a similar cottage industry
   for videos. Also commercial because
   the show ads. Even google search
   doesn't show stack overflow as first
   result anymore.

- Declining Business Model Job Search
   With "low code", the number of programmer
   jobs as we know it will go down. A website
   specialized on job advertising will automatically
   loose revenue:

https://stackoverflow.co/advertising/

Bye

Mild Shock schrieb:

Hi,

Yes exactly, they shoot themselves into the foot.
They adopted the academic closed world attitute,
as can be seen here, not noticing that the John

Doe programmers are not necessarily a theoretical
computer science titanic A-class member:

As an example, consider this question – invented, but not far-fetched: >>> "Random keeps giving me the same numbers. Is it broken?"
This is a low-quality question, in my view.
https://codeblog.jonskeet.uk/2018/03/17/stack-overflow-culture/

Now try the same with ChatGPT:

Q: Random keeps giving me the same numbers. Is it broken?

A: It sounds like your Random object might not be
re-seeded, which could cause it to generate the same
sequence of numbers every time you run your program.
By default, Random is seeded with the current time in
  milliseconds, but if you create it multiple times in
quick succession, it might receive the same seed and
generate the same numbers.

Bye

Mild Shock schrieb:

Hi,

The cest pool of bit rot is finally gone:

(Bit rot, because they did everything in a
facist way ignoring your copyright to prevent
you from deleting you own out dated stuff)

CoincidenceIDontThinkSo
https://9gag.com/gag/abAG4oE

The Fall of Stack Overflow
https://observablehq.com/@ayhanfuat/the-fall-of-stack-overflow

The downfall is related to the appearance of
ChatGPT. I feel pitty for Ulrich Neumerkel false
and others who spammed stackoverflow for years

with their infinite dif/2 nonsense.

Bye

--- Synchronet 3.20a-Linux NewsLink 1.114

From Newsgroup: comp.lang.prolog

Hi,

Strange things happen in the Zig world:

Bun is a JavaScript runtime, package manager,
test runner and bundler built from scratch using
the Zig programming language. It was designed by
Jarred Sumner as a drop-in replacement for Node.js.
Bun uses WebKit's JavaScriptCore as the JavaScript
engine, unlike Node.js and Deno, which both use V8. https://en.wikipedia.org/wiki/Bun_(software)

Should give a try via Dogelog Player for JavaScript.

Python/PyPy (2.288s)
Js/Node (1.595s)
C/clang -O0 (1.566s)
Go (1.528s)
Js/Bun (1.434s)
1 Billion nested loop iterations
https://benjdd.com/loops2/

The above is a micro benchmark, might not give an
indicative whether Bun will choke or not. Lets see.

Bye

Mild Shock schrieb:

Hi,

what is wrong with rust and linux???? https://www.youtube.com/watch?v=T2OKdv4SPdQ

Pure souls are getting purrified to become even purer.

Bye

P.S.: Zig moves to fast for the kernel?
https://ziglang.org/learn/overview/

How about a new Prolog system written in Zip?

Mild Shock schrieb:

Maybe Scryer Prolog is a failure, because Rust
is a failure. Building Scryer Prolog on my machine,
after a git pull of a small change, did just

take this much time:

$ time cargo build --release
[...]
real    6m52.663s
user    7m39.043s
sys     0m3.241s

What did Rust do? And this tabling test case,
2000 shuttle, hangs:

/* Scryer Prolog 0.9.4-107 */
?- use_module(library(tabling)).
    true.
?- [user].
:- table c/1.
c(X) :- c(Y), 0 =< Y, Y < 20000, X is -Y-1.
c(X) :- c(Y), -20000 < Y, Y =< 0, X is -Y+1.
c(0).

?- time(c(_)).
%%% hangs ? %%%

It is solved in a blink in SWI-Prolog:

/* SWI-Prolog 9.3.8 */
?- time(c(_)).
% 540,077 inferences, 0.047 CPU in 0.051 seconds (92% CPU, 11521643 Lips)
true .

The test case is from here:

https://github.com/JanWielemaker/tabling_benchmarks/tree/master

--- Synchronet 3.20a-Linux NewsLink 1.114