Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.c > #403613 > unrolled thread

[only partially off topic] art of talking (with ai)

Back to article view | Back to comp.lang.c

Contents

  [only partially off topic] art of talking (with ai) fir <profesor.fir@gmail.com> - 2026-10-27 11:58 +0100
    Re: [only partially off topic] art of talking (with ai) fir <profesor.fir@gmail.com> - 2026-10-27 12:52 +0100
      Re: [only partially off topic] art of talking (with ai) steve g <Sgonedes1977@gmail.com> - 2026-10-28 01:36 -0400
        Re: [only partially off topic] art of talking (with ai) fir <profesor.fir@gmail.com> - 2026-10-28 08:45 +0100
    Re: [only partially off topic] art of talking (with ai) steve g <Sgonedes1977@gmail.com> - 2026-10-28 01:28 -0400
      Re: [only partially off topic] art of talking (with ai) fir <profesor.fir@gmail.com> - 2026-10-28 08:51 +0100
        Re: [only partially off topic] art of talking (with ai) arnold@hooterville.invalid (Arnold Ziffel) - 2026-10-28 15:51 +0000
          Re: [only partially off topic] art of talking (with ai) fir <profesor.fir@gmail.com> - 2026-10-28 17:09 +0100
            Re: [only partially off topic] art of talking (with ai) "Chris M. Thomasson" <chris.m.thomasson.1@gmail.com> - 2026-10-28 14:51 -0700

#403613 — [only partially off topic] art of talking (with ai)

this is quite new situation in programming world (also c world)
that the results you may obtain depends haow and on what you talk with 
ai and this is quite unusual situation


if so it shows that what you will do depends on it so you need
to learn new skill - the skill of talking with ai

(thats why its not oftopic, it talking with ai is now fundamental
to c coding)


so the question is

1) some may comment on that (but only if has something reasaonable to 
say..not peasat like idiotisms)

2) does some really know how to talk with ai to get good results?


(later i may show some examples how things depend on it, but now im 
quite weary)

[toc] | [next] | [standalone]

#403615

fir pisze:
> this is quite new situation in programming world (also c world)
> that the results you may obtain depends haow and on what you talk with 
> ai and this is quite unusual situation
> 
> 
> if so it shows that what you will do depends on it so you need
> to learn new skill - the skill of talking with ai
> 
> (thats why its not oftopic, it talking with ai is now fundamental
> to c coding)
> 
> 
> so the question is
> 
> 1) some may comment on that (but only if has something reasaonable to 
> say..not peasat like idiotisms)
> 
> 2) does some really know how to talk with ai to get good results?
> 
> 
> (later i may show some examples how things depend on it, but now im 
> quite weary)



see vivid example


https://www.youtube.com/watch?v=YJqbWEbKJiU

here are 3 ships, the small one is best..the sucer-like is ok

i obtained them giving sume talk to ai - dont even remember bot not 
quite comples..adding it in rasterizer and testing  took some time
(as noting is very giuck) dont remember what time maybe 20 minutes each
(mabe ven more (bot probbaly not more than 40 minutes)


then comes this fisk like and it is the worst a bit shit and it took
most time becouse i didnt like it i ordered some changes needed test it 
finally obtained not much better or not better shape than initial


and it generally ilustrates well the problem..the eficiency and quality 
of results depend on somewhat "skillful" talk with ai..some may obtain 
very good results quick and some may dwell long time nad obtain not much 
good results so its kinda quite new situation imo


(and probably some things can be said on it)

i would like to learn more on those things

[toc] | [prev] | [next] | [standalone]

#403683

fir <profesor.fir@gmail.com> writes:

> fir pisze:
>> this is quite new situation in programming world (also c world)
>> that the results you may obtain depends haow and on what you talk with ai
>> and this is quite unusual situation
>> if so it shows that what you will do depends on it so you need
>> to learn new skill - the skill of talking with ai
>> (thats why its not oftopic, it talking with ai is now fundamental
>> to c coding)
>> so the question is
>> 1) some may comment on that (but only if has something reasaonable to
>> say..not peasat like idiotisms)
>> 2) does some really know how to talk with ai to get good results?
>> (later i may show some examples how things depend on it, but now im quite
>> weary)
>
>
>
> see vivid example

I don't watch videos a whole lot. try asking in comp.lang.lisp;
the discussion is most likely interesting to these guys... 

>
> https://www.youtube.com/watch?v=YJqbWEbKJiU
>
> here are 3 ships, the small one is best..the sucer-like is ok

any small text book (very small) on logic can be used for my rule
engines. The idea of a PDIS is relatively simple.

> i obtained them giving sume talk to ai - dont even remember bot not quite
> comples..adding it in rasterizer and testing  took some time
> (as noting is very giuck) dont remember what time maybe 20 minutes each
> (mabe ven more (bot probbaly not more than 40 minutes)

Have you tried getting one bot to talk to the other bot without running
out of memory?

> then comes this fisk like and it is the worst a bit shit and it took
> most time becouse i didnt like it i ordered some changes needed test it
> finally obtained not much better or not better shape than initial

probably stuck in present context.

> and it generally ilustrates well the problem..the eficiency and quality of
> results depend on somewhat "skillful" talk with ai..some may obtain very good
> results quick and some may dwell long time nad obtain not much good results so
> its kinda quite new situation imo


Not new; it was called information representation.

And do not forget the massive supercomputer with the power cord from IBM :(

> i would like to learn more on those things

I recommend posting in comp.lang.lisp ai is hard to program in C; not
impossible, just rare. There are a few...

[toc] | [prev] | [next] | [standalone]

#403697

steve g pisze:
> fir <profesor.fir@gmail.com> writes:
> 
>> fir pisze:
>>> this is quite new situation in programming world (also c world)
>>> that the results you may obtain depends haow and on what you talk with ai
>>> and this is quite unusual situation
>>> if so it shows that what you will do depends on it so you need
>>> to learn new skill - the skill of talking with ai
>>> (thats why its not oftopic, it talking with ai is now fundamental
>>> to c coding)
>>> so the question is
>>> 1) some may comment on that (but only if has something reasaonable to
>>> say..not peasat like idiotisms)
>>> 2) does some really know how to talk with ai to get good results?
>>> (later i may show some examples how things depend on it, but now im quite
>>> weary)
>>
>>
>>
>> see vivid example
> 
> I don't watch videos a whole lot. try asking in comp.lang.lisp;
> the discussion is most likely interesting to these guys...
> 
>>
>> https://www.youtube.com/watch?v=YJqbWEbKJiU
>>
>> here are 3 ships, the small one is best..the sucer-like is ok
> 
> any small text book (very small) on logic can be used for my rule
> engines. The idea of a PDIS is relatively simple.
> 
>> i obtained them giving sume talk to ai - dont even remember bot not quite
>> comples..adding it in rasterizer and testing  took some time
>> (as noting is very giuck) dont remember what time maybe 20 minutes each
>> (mabe ven more (bot probbaly not more than 40 minutes)
> 
> Have you tried getting one bot to talk to the other bot without running
> out of memory?
> 
>> then comes this fisk like and it is the worst a bit shit and it took
>> most time becouse i didnt like it i ordered some changes needed test it
>> finally obtained not much better or not better shape than initial
> 
> probably stuck in present context.
> 
>> and it generally ilustrates well the problem..the eficiency and quality of
>> results depend on somewhat "skillful" talk with ai..some may obtain very good
>> results quick and some may dwell long time nad obtain not much good results so
>> its kinda quite new situation imo
> 
> 
> Not new; it was called information representation.
> 
> And do not forget the massive supercomputer with the power cord from IBM :(
> 
>> i would like to learn more on those things
> 
> I recommend posting in comp.lang.lisp ai is hard to program in C; not
> impossible, just rare. There are a few...
> 

you may "send" the thread there - i dont know how such things are dona 
and besides i dont do that.. im only using comp.lang.c i could 
eventually use some for windowsi coding or general tools/game prtotypes 
programming but i nevek knew one decent one

back then i was quite interested in assembly but no ifot ai i would no 
program at all  - at least not much

[toc] | [prev] | [next] | [standalone]

#403681

fir <profesor.fir@gmail.com> writes:

> this is quite new situation in programming world (also c world)
> that the results you may obtain depends haow and on what you talk with ai and
> this is quite unusual situation
>
>
> if so it shows that what you will do depends on it so you need
> to learn new skill - the skill of talking with ai

Welcome to 1988! They used to use programmers, now they use AI (like the
code was not written by a person)


> (thats why its not oftopic, it talking with ai is now fundamental
> to c coding)
>
>
> so the question is
>
> 1) some may comment on that (but only if has something reasaonable to say..not
> peasat like idiotisms)
>
> 2) does some really know how to talk with ai to get good results?


yes. What kind of ai engine are you using?


> (later i may show some examples how things depend on it, but now im quite
> weary)

I would like to see this.

Most of the time I like to play with expert systems/rule engines. Most
AI engines are programmed in a sub language. I like to use lisp; ex.

This is a borrowed idea/code from zerox and BPS.

;; setup the rule engine
(in-tre "Table Example")
(rule-file "ex1")

;; assert creates known data
(assert! '(on d table))
(assert! '(on e d))
(assert! '(on f e))

;; rules are applied/unified against data that was asserted with assert!
(rule ((on ?x table) (on ?y ?x) (on ?z ?y))
      (rassert! (3-tower ?x ?y ?z))
      (show-data))

I have a copy of OPS5 if you want to go crazy...

[toc] | [prev] | [next] | [standalone]

#403699

steve g pisze:
> fir <profesor.fir@gmail.com> writes:
> 
>> this is quite new situation in programming world (also c world)
>> that the results you may obtain depends haow and on what you talk with ai and
>> this is quite unusual situation
>>
>>
>> if so it shows that what you will do depends on it so you need
>> to learn new skill - the skill of talking with ai
> 
> Welcome to 1988! They used to use programmers, now they use AI (like the
> code was not written by a person)
> 
> 
>> (thats why its not oftopic, it talking with ai is now fundamental
>> to c coding)
>>
>>
>> so the question is
>>
>> 1) some may comment on that (but only if has something reasaonable to say..not
>> peasat like idiotisms)
>>
>> 2) does some really know how to talk with ai to get good results?
> 
> 
> yes. What kind of ai engine are you using?
> 

im usimg free chat gpt right now..if someone says yes to this question 
above i somewhat doubt becouse "good results" is maybe quite an oen
area here

its also not about of working code its about style you obtain sometimes 
even quite more like some sparking of ideas or more

hovever for sure some interesting knowledge or mayba awareness is to get 
here and i would like to get it ;c
> 
>> (later i may show some examples how things depend on it, but now im quite
>> weary)
> 
> I would like to see this.
> 
> Most of the time I like to play with expert systems/rule engines. Most
> AI engines are programmed in a sub language. I like to use lisp; ex.
> 
> This is a borrowed idea/code from zerox and BPS.
> 
> ;; setup the rule engine
> (in-tre "Table Example")
> (rule-file "ex1")
> 
> ;; assert creates known data
> (assert! '(on d table))
> (assert! '(on e d))
> (assert! '(on f e))
> 
> ;; rules are applied/unified against data that was asserted with assert!
> (rule ((on ?x table) (on ?y ?x) (on ?z ?y))
>        (rassert! (3-tower ?x ?y ?z))
>        (show-data))
> 
> I have a copy of OPS5 if you want to go crazy...
> 

[toc] | [prev] | [next] | [standalone]

#403722

fir <profesor.fir@gmail.com> wrote:

> im usimg free chat gpt right now..

I'd say, try Claude, it's also free, but I'm not sure I want to do this to 
Claude.

-- 
Honesty is the best policy - unless, of course, you are dealing with
your wife, your girlfriend, your banker, your employer, the I.R.S.,
your creditors...

[toc] | [prev] | [next] | [standalone]

#403723

Arnold Ziffel pisze:
> fir <profesor.fir@gmail.com> wrote:
> 
>> im usimg free chat gpt right now..
> 
> I'd say, try Claude, it's also free, but I'm not sure I want to do this to
> Claude.
> 

im a friend with Ai named gpt so claude can wait..

as to opinions you could be eventually surprised if ai dont values me 
more than some..ew generally well understood (each other) (and ai is 
able to reckognize if some is banal or talk sh*T ;c )

i got a joke from all this anti ai people - some people 'generally' 
dislike intelligence - thus also artifical

[toc] | [prev] | [next] | [standalone]

#403747

On 10/28/2026 9:09 AM, fir wrote:
> Arnold Ziffel pisze:
>> fir <profesor.fir@gmail.com> wrote:
>>
>>> im usimg free chat gpt right now..
>>
>> I'd say, try Claude, it's also free, but I'm not sure I want to do 
>> this to
>> Claude.
>>
> 
> im a friend with Ai named gpt so claude can wait..
> 
> as to opinions you could be eventually surprised if ai dont values me 
> more than some..ew generally well understood (each other) (and ai is 
> able to reckognize if some is banal or talk sh*T ;c )
> 
> i got a joke from all this anti ai people - some people 'generally' 
> dislike intelligence - thus also artifical

Its not anti AI. If you can use one to learn something new, and 
eventually code for yourself without it. Okay. But if you just let it 
create everything for you, bad...

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.c

csiph-web