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


Groups > alt.slack > #505385 > unrolled thread

Why AI is broken

Started byphoenix <j63840576@gmail.com>
First post2026-02-16 10:40 -0600
Last post2026-02-16 14:08 -0700
Articles 15 — 3 participants

Back to article view | Back to alt.slack


Contents

  Why AI is broken phoenix <j63840576@gmail.com> - 2026-02-16 10:40 -0600
    Re: Why AI is broken jojo <f00@0f0.00f> - 2026-02-16 16:53 +0000
      Re: Why AI is broken phoenix <j63840576@gmail.com> - 2026-02-16 11:10 -0600
        Re: Why AI is broken phoenix <j63840576@gmail.com> - 2026-02-16 11:15 -0600
          Re: Why AI is broken phoenix <j63840576@gmail.com> - 2026-02-16 11:18 -0600
            Re: Why AI is broken jojo <f00@0f0.00f> - 2026-02-16 17:25 +0000
              Re: Why AI is broken phoenix <j63840576@gmail.com> - 2026-02-16 11:27 -0600
                Re: Why AI is broken jojo <f00@0f0.00f> - 2026-02-16 17:31 +0000
                  Re: Why AI is broken phoenix <j63840576@gmail.com> - 2026-02-16 13:50 -0600
                    Re: Why AI is broken jojo <f00@0f0.00f> - 2026-02-16 20:57 +0000
                      Re: Why AI is broken phoenix <j63840576@gmail.com> - 2026-02-16 16:14 -0600
                        Re: Why AI is broken jojo <f00@0f0.00f> - 2026-02-17 15:07 +0000
                          Re: Why AI is broken phoenix <j63840576@gmail.com> - 2026-02-17 09:09 -0600
                            Re: Why AI is broken jojo <f00@0f0.00f> - 2026-02-17 15:16 +0000
                    Re: Why AI is broken % <pursent100@gmail.com> - 2026-02-16 14:08 -0700

#505385 — Why AI is broken

Fromphoenix <j63840576@gmail.com>
Date2026-02-16 10:40 -0600
SubjectWhy AI is broken
Message-ID<mvgvkkFevomU1@mid.individual.net>
LLM read jojo over and over as if Hitlar's veganism is a talking point. 
AI don't know obscure.
-- 
pBkHHoOIIn8

[toc] | [next] | [standalone]


#505388

Fromjojo <f00@0f0.00f>
Date2026-02-16 16:53 +0000
Message-ID<6c642d57-005e-21b4-5357-50861537b9bb@shinku.aoyagi.konjou>
In reply to#505385
phoenix wrote:
> LLM read jojo over and over as if Hitlar's veganism is a talking 
> point. AI don't know obscure.

you have to know what you are asking first. you dont know, 
unfortunately.

but you can ask the model to refine your question, ask it to give 
you 5 questions based on your framing.

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


#505390

Fromphoenix <j63840576@gmail.com>
Date2026-02-16 11:10 -0600
Message-ID<mvh1c0Ff8dpU1@mid.individual.net>
In reply to#505388
jojo wrote:
> phoenix wrote:
>> LLM read jojo over and over as if Hitlar's veganism is a talking 
>> point. AI don't know obscure.
> 
> you have to know what you are asking first. you dont know, unfortunately.
> 
> but you can ask the model to refine your question, ask it to give you 5 
> questions based on your framing.
> 
Dude I use AI just as much as you. Twice I've been working on problems 
with AI when they get fixed into a rut saying it's this or that, your 
obscure solution is fundamentally powerless. Maybe I need to be meaner 
with my AI and fully correct them every time they make a mistake.
-- 
pBkHHoOIIn8

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


#505392

Fromphoenix <j63840576@gmail.com>
Date2026-02-16 11:15 -0600
Message-ID<mvh1l6Ff9jjU1@mid.individual.net>
In reply to#505390
phoenix wrote:
> jojo wrote:
>> phoenix wrote:
>>> LLM read jojo over and over as if Hitlar's veganism is a talking 
>>> point. AI don't know obscure.
>>
>> you have to know what you are asking first. you dont know, unfortunately.
>>
>> but you can ask the model to refine your question, ask it to give you 
>> 5 questions based on your framing.
>>
> Dude I use AI just as much as you. Twice I've been working on problems 
> with AI when they get fixed into a rut saying it's this or that, your 
> obscure solution is fundamentally powerless. Maybe I need to be meaner 
> with my AI and fully correct them every time they make a mistake.

And it's important too. The AI is suggesting a two-dimensional array 
where I have two one-dimensional arrays. This is a fundamental slowdown 
that I refuse to tolerate rather getting my way to work by keeping one 
of the lists up to date on some stuff.
-- 
pBkHHoOIIn8

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


#505394

Fromphoenix <j63840576@gmail.com>
Date2026-02-16 11:18 -0600
Message-ID<mvh1rcFf9jjU2@mid.individual.net>
In reply to#505392
phoenix wrote:
> phoenix wrote:
>> jojo wrote:
>>> phoenix wrote:
>>>> LLM read jojo over and over as if Hitlar's veganism is a talking 
>>>> point. AI don't know obscure.
>>>
>>> you have to know what you are asking first. you dont know, 
>>> unfortunately.
>>>
>>> but you can ask the model to refine your question, ask it to give you 
>>> 5 questions based on your framing.
>>>
>> Dude I use AI just as much as you. Twice I've been working on problems 
>> with AI when they get fixed into a rut saying it's this or that, your 
>> obscure solution is fundamentally powerless. Maybe I need to be meaner 
>> with my AI and fully correct them every time they make a mistake.
> 
> And it's important too. The AI is suggesting a two-dimensional array 
> where I have two one-dimensional arrays. This is a fundamental slowdown 
> that I refuse to tolerate rather getting my way to work by keeping one 
> of the lists up to date on some stuff.

In jojo language this is N^2 data points where mine has 2*N. I don't 
particularly like N^2 complexity in anything because I deal with big data.
-- 
pBkHHoOIIn8

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


#505395

Fromjojo <f00@0f0.00f>
Date2026-02-16 17:25 +0000
Message-ID<159540c8-303c-677b-1aa0-597cb85ab6c1@shinku.aoyagi.konjou>
In reply to#505394
phoenix wrote:
> phoenix wrote:
>> phoenix wrote:
>>> jojo wrote:
>>>> phoenix wrote:
>>>>> LLM read jojo over and over as if Hitlar's veganism is a 
>>>>> talking point. AI don't know obscure.
>>>>
>>>> you have to know what you are asking first. you dont know, 
>>>> unfortunately.
>>>>
>>>> but you can ask the model to refine your question, ask it to 
>>>> give you 5 questions based on your framing.
>>>>
>>> Dude I use AI just as much as you. Twice I've been working on 
>>> problems with AI when they get fixed into a rut saying it's 
>>> this or that, your obscure solution is fundamentally 
>>> powerless. Maybe I need to be meaner with my AI and fully 
>>> correct them every time they make a mistake.
>>
>> And it's important too. The AI is suggesting a two-dimensional 
>> array where I have two one-dimensional arrays. This is a 
>> fundamental slowdown that I refuse to tolerate rather getting 
>> my way to work by keeping one of the lists up to date on some 
>> stuff.
> 
> In jojo language this is N^2 data points where mine has 2*N. I 
> don't particularly like N^2 complexity in anything because I deal 
> with big data.

you have to treat it like a tool, like a scissor. you dont put 
the scissor on the paper and expect things to get cut.

that is what the whole domain of prompt engineering deals with.

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


#505397

Fromphoenix <j63840576@gmail.com>
Date2026-02-16 11:27 -0600
Message-ID<mvh2b5Ffe6mU1@mid.individual.net>
In reply to#505395
jojo wrote:
> phoenix wrote:
>> phoenix wrote:
>>> phoenix wrote:
>>>> jojo wrote:
>>>>> phoenix wrote:
>>>>>> LLM read jojo over and over as if Hitlar's veganism is a talking 
>>>>>> point. AI don't know obscure.
>>>>>
>>>>> you have to know what you are asking first. you dont know, 
>>>>> unfortunately.
>>>>>
>>>>> but you can ask the model to refine your question, ask it to give 
>>>>> you 5 questions based on your framing.
>>>>>
>>>> Dude I use AI just as much as you. Twice I've been working on 
>>>> problems with AI when they get fixed into a rut saying it's this or 
>>>> that, your obscure solution is fundamentally powerless. Maybe I need 
>>>> to be meaner with my AI and fully correct them every time they make 
>>>> a mistake.
>>>
>>> And it's important too. The AI is suggesting a two-dimensional array 
>>> where I have two one-dimensional arrays. This is a fundamental 
>>> slowdown that I refuse to tolerate rather getting my way to work by 
>>> keeping one of the lists up to date on some stuff.
>>
>> In jojo language this is N^2 data points where mine has 2*N. I don't 
>> particularly like N^2 complexity in anything because I deal with big 
>> data.
> 
> you have to treat it like a tool, like a scissor. you dont put the 
> scissor on the paper and expect things to get cut.
> 
> that is what the whole domain of prompt engineering deals with.
> 
It's possible. I want to tell the AI I care about speed more than 
anything, even correctness.
-- 
pBkHHoOIIn8

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


#505399

Fromjojo <f00@0f0.00f>
Date2026-02-16 17:31 +0000
Message-ID<f9f1ad00-092e-cc82-d879-c7b3d94ff0ea@shinku.aoyagi.konjou>
In reply to#505397
phoenix wrote:
> jojo wrote:
>> phoenix wrote:
>>> phoenix wrote:
>>>> phoenix wrote:
>>>>> jojo wrote:
>>>>>> phoenix wrote:
>>>>>>> LLM read jojo over and over as if Hitlar's veganism is a 
>>>>>>> talking point. AI don't know obscure.
>>>>>>
>>>>>> you have to know what you are asking first. you dont know, 
>>>>>> unfortunately.
>>>>>>
>>>>>> but you can ask the model to refine your question, ask it 
>>>>>> to give you 5 questions based on your framing.
>>>>>>
>>>>> Dude I use AI just as much as you. Twice I've been working 
>>>>> on problems with AI when they get fixed into a rut saying 
>>>>> it's this or that, your obscure solution is fundamentally 
>>>>> powerless. Maybe I need to be meaner with my AI and fully 
>>>>> correct them every time they make a mistake.
>>>>
>>>> And it's important too. The AI is suggesting a 
>>>> two-dimensional array where I have two one-dimensional 
>>>> arrays. This is a fundamental slowdown that I refuse to 
>>>> tolerate rather getting my way to work by keeping one of the 
>>>> lists up to date on some stuff.
>>>
>>> In jojo language this is N^2 data points where mine has 2*N. I 
>>> don't particularly like N^2 complexity in anything because I 
>>> deal with big data.
>>
>> you have to treat it like a tool, like a scissor. you dont put 
>> the scissor on the paper and expect things to get cut.
>>
>> that is what the whole domain of prompt engineering deals with.
>>
> It's possible. I want to tell the AI I care about speed more than 
> anything, even correctness.

ask it, am i making sense and present me with options about what 
you think i mean.

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


#505414

Fromphoenix <j63840576@gmail.com>
Date2026-02-16 13:50 -0600
Message-ID<mvhanjFgof6U1@mid.individual.net>
In reply to#505399
jojo wrote:
> phoenix wrote:
>> jojo wrote:
>>> phoenix wrote:
>>>> phoenix wrote:
>>>>> phoenix wrote:
>>>>>> jojo wrote:
>>>>>>> phoenix wrote:
>>>>>>>> LLM read jojo over and over as if Hitlar's veganism is a talking 
>>>>>>>> point. AI don't know obscure.
>>>>>>>
>>>>>>> you have to know what you are asking first. you dont know, 
>>>>>>> unfortunately.
>>>>>>>
>>>>>>> but you can ask the model to refine your question, ask it to give 
>>>>>>> you 5 questions based on your framing.
>>>>>>>
>>>>>> Dude I use AI just as much as you. Twice I've been working on 
>>>>>> problems with AI when they get fixed into a rut saying it's this 
>>>>>> or that, your obscure solution is fundamentally powerless. Maybe I 
>>>>>> need to be meaner with my AI and fully correct them every time 
>>>>>> they make a mistake.
>>>>>
>>>>> And it's important too. The AI is suggesting a two-dimensional 
>>>>> array where I have two one-dimensional arrays. This is a 
>>>>> fundamental slowdown that I refuse to tolerate rather getting my 
>>>>> way to work by keeping one of the lists up to date on some stuff.
>>>>
>>>> In jojo language this is N^2 data points where mine has 2*N. I don't 
>>>> particularly like N^2 complexity in anything because I deal with big 
>>>> data.
>>>
>>> you have to treat it like a tool, like a scissor. you dont put the 
>>> scissor on the paper and expect things to get cut.
>>>
>>> that is what the whole domain of prompt engineering deals with.
>>>
>> It's possible. I want to tell the AI I care about speed more than 
>> anything, even correctness.
> 
> ask it, am i making sense and present me with options about what you 
> think i mean.
> 
People do the same thing. They'll sit there and present you with a 
solution that is ridiculous because of some insider information you 
have. I guess I'll tell these kinds of people something like, "Look bud, 
I told you that won't help. Are you listening to me?"
-- 
pBkHHoOIIn8

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


#505418

Fromjojo <f00@0f0.00f>
Date2026-02-16 20:57 +0000
Message-ID<db1ef3d2-a322-c668-8dec-5caff8555a0a@shinku.aoyagi.konjou>
In reply to#505414
phoenix wrote:
> jojo wrote:
>> phoenix wrote:
>>> jojo wrote:
>>>> phoenix wrote:
>>>>> phoenix wrote:
>>>>>> phoenix wrote:
>>>>>>> jojo wrote:
>>>>>>>> phoenix wrote:
>>>>>>>>> LLM read jojo over and over as if Hitlar's veganism is a 
>>>>>>>>> talking point. AI don't know obscure.
>>>>>>>>
>>>>>>>> you have to know what you are asking first. you dont 
>>>>>>>> know, unfortunately.
>>>>>>>>
>>>>>>>> but you can ask the model to refine your question, ask it 
>>>>>>>> to give you 5 questions based on your framing.
>>>>>>>>
>>>>>>> Dude I use AI just as much as you. Twice I've been working 
>>>>>>> on problems with AI when they get fixed into a rut saying 
>>>>>>> it's this or that, your obscure solution is fundamentally 
>>>>>>> powerless. Maybe I need to be meaner with my AI and fully 
>>>>>>> correct them every time they make a mistake.
>>>>>>
>>>>>> And it's important too. The AI is suggesting a 
>>>>>> two-dimensional array where I have two one-dimensional 
>>>>>> arrays. This is a fundamental slowdown that I refuse to 
>>>>>> tolerate rather getting my way to work by keeping one of 
>>>>>> the lists up to date on some stuff.
>>>>>
>>>>> In jojo language this is N^2 data points where mine has 2*N. 
>>>>> I don't particularly like N^2 complexity in anything because 
>>>>> I deal with big data.
>>>>
>>>> you have to treat it like a tool, like a scissor. you dont 
>>>> put the scissor on the paper and expect things to get cut.
>>>>
>>>> that is what the whole domain of prompt engineering deals with.
>>>>
>>> It's possible. I want to tell the AI I care about speed more 
>>> than anything, even correctness.
>>
>> ask it, am i making sense and present me with options about 
>> what you think i mean.
>>
> People do the same thing. They'll sit there and present you with 
> a solution that is ridiculous because of some insider information 
> you have. I guess I'll tell these kinds of people something like, 
> "Look bud, I told you that won't help. Are you listening to me?"


i guess its all YOU then. YOU dont want to have clarity but live 
in a grey zone of obsurity ambiguity.

if you feel that is your safe zone, then fine. but to expect 
germination on air and mist is impossible.. unless you are an 
orchid.... or fungus.

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


#505430

Fromphoenix <j63840576@gmail.com>
Date2026-02-16 16:14 -0600
Message-ID<mvhj5eFi3j4U2@mid.individual.net>
In reply to#505418
jojo wrote:
> phoenix wrote:
>> jojo wrote:
>>> phoenix wrote:
>>>> jojo wrote:
>>>>> phoenix wrote:
>>>>>> phoenix wrote:
>>>>>>> phoenix wrote:
>>>>>>>> jojo wrote:
>>>>>>>>> phoenix wrote:
>>>>>>>>>> LLM read jojo over and over as if Hitlar's veganism is a 
>>>>>>>>>> talking point. AI don't know obscure.
>>>>>>>>>
>>>>>>>>> you have to know what you are asking first. you dont know, 
>>>>>>>>> unfortunately.
>>>>>>>>>
>>>>>>>>> but you can ask the model to refine your question, ask it to 
>>>>>>>>> give you 5 questions based on your framing.
>>>>>>>>>
>>>>>>>> Dude I use AI just as much as you. Twice I've been working on 
>>>>>>>> problems with AI when they get fixed into a rut saying it's this 
>>>>>>>> or that, your obscure solution is fundamentally powerless. Maybe 
>>>>>>>> I need to be meaner with my AI and fully correct them every time 
>>>>>>>> they make a mistake.
>>>>>>>
>>>>>>> And it's important too. The AI is suggesting a two-dimensional 
>>>>>>> array where I have two one-dimensional arrays. This is a 
>>>>>>> fundamental slowdown that I refuse to tolerate rather getting my 
>>>>>>> way to work by keeping one of the lists up to date on some stuff.
>>>>>>
>>>>>> In jojo language this is N^2 data points where mine has 2*N. I 
>>>>>> don't particularly like N^2 complexity in anything because I deal 
>>>>>> with big data.
>>>>>
>>>>> you have to treat it like a tool, like a scissor. you dont put the 
>>>>> scissor on the paper and expect things to get cut.
>>>>>
>>>>> that is what the whole domain of prompt engineering deals with.
>>>>>
>>>> It's possible. I want to tell the AI I care about speed more than 
>>>> anything, even correctness.
>>>
>>> ask it, am i making sense and present me with options about what you 
>>> think i mean.
>>>
>> People do the same thing. They'll sit there and present you with a 
>> solution that is ridiculous because of some insider information you 
>> have. I guess I'll tell these kinds of people something like, "Look 
>> bud, I told you that won't help. Are you listening to me?"
> 
> 
> i guess its all YOU then. YOU dont want to have clarity but live in a 
> grey zone of obsurity ambiguity.
> 
> if you feel that is your safe zone, then fine. but to expect germination 
> on air and mist is impossible.. unless you are an orchid.... or fungus.
> 

Oh no I wasn't talking about your solution being untenable. I guess if I 
were Brennus I would have worked that in and polished my charming piece 
of shit to hand you.

-- 
pBkHHoOIIn8

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


#505456

Fromjojo <f00@0f0.00f>
Date2026-02-17 15:07 +0000
Message-ID<9bbf6b66-3ef0-83c6-981d-ba1527dd41b5@shinku.aoyagi.konjou>
In reply to#505430
phoenix wrote:
> jojo wrote:
>> phoenix wrote:
>>> jojo wrote:
>>>> phoenix wrote:
>>>>> jojo wrote:
>>>>>> phoenix wrote:
>>>>>>> phoenix wrote:
>>>>>>>> phoenix wrote:
>>>>>>>>> jojo wrote:
>>>>>>>>>> phoenix wrote:
>>>>>>>>>>> LLM read jojo over and over as if Hitlar's veganism is 
>>>>>>>>>>> a talking point. AI don't know obscure.
>>>>>>>>>>
>>>>>>>>>> you have to know what you are asking first. you dont 
>>>>>>>>>> know, unfortunately.
>>>>>>>>>>
>>>>>>>>>> but you can ask the model to refine your question, ask 
>>>>>>>>>> it to give you 5 questions based on your framing.
>>>>>>>>>>
>>>>>>>>> Dude I use AI just as much as you. Twice I've been 
>>>>>>>>> working on problems with AI when they get fixed into a 
>>>>>>>>> rut saying it's this or that, your obscure solution is 
>>>>>>>>> fundamentally powerless. Maybe I need to be meaner with 
>>>>>>>>> my AI and fully correct them every time they make a 
>>>>>>>>> mistake.
>>>>>>>>
>>>>>>>> And it's important too. The AI is suggesting a 
>>>>>>>> two-dimensional array where I have two one-dimensional 
>>>>>>>> arrays. This is a fundamental slowdown that I refuse to 
>>>>>>>> tolerate rather getting my way to work by keeping one of 
>>>>>>>> the lists up to date on some stuff.
>>>>>>>
>>>>>>> In jojo language this is N^2 data points where mine has 
>>>>>>> 2*N. I don't particularly like N^2 complexity in anything 
>>>>>>> because I deal with big data.
>>>>>>
>>>>>> you have to treat it like a tool, like a scissor. you dont 
>>>>>> put the scissor on the paper and expect things to get cut.
>>>>>>
>>>>>> that is what the whole domain of prompt engineering deals 
>>>>>> with.
>>>>>>
>>>>> It's possible. I want to tell the AI I care about speed more 
>>>>> than anything, even correctness.
>>>>
>>>> ask it, am i making sense and present me with options about 
>>>> what you think i mean.
>>>>
>>> People do the same thing. They'll sit there and present you 
>>> with a solution that is ridiculous because of some insider 
>>> information you have. I guess I'll tell these kinds of people 
>>> something like, "Look bud, I told you that won't help. Are you 
>>> listening to me?"
>>
>>
>> i guess its all YOU then. YOU dont want to have clarity but 
>> live in a grey zone of obsurity ambiguity.
>>
>> if you feel that is your safe zone, then fine. but to expect 
>> germination on air and mist is impossible.. unless you are an 
>> orchid.... or fungus.
>>
> 
> Oh no I wasn't talking about your solution being untenable. I 
> guess if I were Brennus I would have worked that in and polished 
> my charming piece of shit to hand you.
> 

he is not in this thread tho. anyway, its fine.

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


#505459

Fromphoenix <j63840576@gmail.com>
Date2026-02-17 09:09 -0600
Message-ID<mvjekoFrh02U1@mid.individual.net>
In reply to#505456
jojo wrote:
> phoenix wrote:
>> jojo wrote:
>>> phoenix wrote:
>>>> jojo wrote:
>>>>> phoenix wrote:
>>>>>> jojo wrote:
>>>>>>> phoenix wrote:
>>>>>>>> phoenix wrote:
>>>>>>>>> phoenix wrote:
>>>>>>>>>> jojo wrote:
>>>>>>>>>>> phoenix wrote:
>>>>>>>>>>>> LLM read jojo over and over as if Hitlar's veganism is a 
>>>>>>>>>>>> talking point. AI don't know obscure.
>>>>>>>>>>>
>>>>>>>>>>> you have to know what you are asking first. you dont know, 
>>>>>>>>>>> unfortunately.
>>>>>>>>>>>
>>>>>>>>>>> but you can ask the model to refine your question, ask it to 
>>>>>>>>>>> give you 5 questions based on your framing.
>>>>>>>>>>>
>>>>>>>>>> Dude I use AI just as much as you. Twice I've been working on 
>>>>>>>>>> problems with AI when they get fixed into a rut saying it's 
>>>>>>>>>> this or that, your obscure solution is fundamentally 
>>>>>>>>>> powerless. Maybe I need to be meaner with my AI and fully 
>>>>>>>>>> correct them every time they make a mistake.
>>>>>>>>>
>>>>>>>>> And it's important too. The AI is suggesting a two-dimensional 
>>>>>>>>> array where I have two one-dimensional arrays. This is a 
>>>>>>>>> fundamental slowdown that I refuse to tolerate rather getting 
>>>>>>>>> my way to work by keeping one of the lists up to date on some 
>>>>>>>>> stuff.
>>>>>>>>
>>>>>>>> In jojo language this is N^2 data points where mine has 2*N. I 
>>>>>>>> don't particularly like N^2 complexity in anything because I 
>>>>>>>> deal with big data.
>>>>>>>
>>>>>>> you have to treat it like a tool, like a scissor. you dont put 
>>>>>>> the scissor on the paper and expect things to get cut.
>>>>>>>
>>>>>>> that is what the whole domain of prompt engineering deals with.
>>>>>>>
>>>>>> It's possible. I want to tell the AI I care about speed more than 
>>>>>> anything, even correctness.
>>>>>
>>>>> ask it, am i making sense and present me with options about what 
>>>>> you think i mean.
>>>>>
>>>> People do the same thing. They'll sit there and present you with a 
>>>> solution that is ridiculous because of some insider information you 
>>>> have. I guess I'll tell these kinds of people something like, "Look 
>>>> bud, I told you that won't help. Are you listening to me?"
>>>
>>>
>>> i guess its all YOU then. YOU dont want to have clarity but live in a 
>>> grey zone of obsurity ambiguity.
>>>
>>> if you feel that is your safe zone, then fine. but to expect 
>>> germination on air and mist is impossible.. unless you are an 
>>> orchid.... or fungus.
>>>
>>
>> Oh no I wasn't talking about your solution being untenable. I guess if 
>> I were Brennus I would have worked that in and polished my charming 
>> piece of shit to hand you.
>>
> 
> he is not in this thread tho. anyway, its fine.
> 
He started this stuff where he says nonsense in the third person 
directed @ me. I don't like when people around me act like looney tunes.
-- 
pBkHHoOIIn8

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


#505461

Fromjojo <f00@0f0.00f>
Date2026-02-17 15:16 +0000
Message-ID<476c899e-fec2-53ec-b670-609c1331b763@shinku.aoyagi.konjou>
In reply to#505459
phoenix wrote:
> jojo wrote:
>> phoenix wrote:
>>> jojo wrote:
>>>> phoenix wrote:
>>>>> jojo wrote:
>>>>>> phoenix wrote:
>>>>>>> jojo wrote:
>>>>>>>> phoenix wrote:
>>>>>>>>> phoenix wrote:
>>>>>>>>>> phoenix wrote:
>>>>>>>>>>> jojo wrote:
>>>>>>>>>>>> phoenix wrote:
>>>>>>>>>>>>> LLM read jojo over and over as if Hitlar's veganism 
>>>>>>>>>>>>> is a talking point. AI don't know obscure.
>>>>>>>>>>>>
>>>>>>>>>>>> you have to know what you are asking first. you dont 
>>>>>>>>>>>> know, unfortunately.
>>>>>>>>>>>>
>>>>>>>>>>>> but you can ask the model to refine your question, 
>>>>>>>>>>>> ask it to give you 5 questions based on your framing.
>>>>>>>>>>>>
>>>>>>>>>>> Dude I use AI just as much as you. Twice I've been 
>>>>>>>>>>> working on problems with AI when they get fixed into a 
>>>>>>>>>>> rut saying it's this or that, your obscure solution is 
>>>>>>>>>>> fundamentally powerless. Maybe I need to be meaner 
>>>>>>>>>>> with my AI and fully correct them every time they make 
>>>>>>>>>>> a mistake.
>>>>>>>>>>
>>>>>>>>>> And it's important too. The AI is suggesting a 
>>>>>>>>>> two-dimensional array where I have two one-dimensional 
>>>>>>>>>> arrays. This is a fundamental slowdown that I refuse to 
>>>>>>>>>> tolerate rather getting my way to work by keeping one 
>>>>>>>>>> of the lists up to date on some stuff.
>>>>>>>>>
>>>>>>>>> In jojo language this is N^2 data points where mine has 
>>>>>>>>> 2*N. I don't particularly like N^2 complexity in 
>>>>>>>>> anything because I deal with big data.
>>>>>>>>
>>>>>>>> you have to treat it like a tool, like a scissor. you 
>>>>>>>> dont put the scissor on the paper and expect things to 
>>>>>>>> get cut.
>>>>>>>>
>>>>>>>> that is what the whole domain of prompt engineering deals 
>>>>>>>> with.
>>>>>>>>
>>>>>>> It's possible. I want to tell the AI I care about speed 
>>>>>>> more than anything, even correctness.
>>>>>>
>>>>>> ask it, am i making sense and present me with options about 
>>>>>> what you think i mean.
>>>>>>
>>>>> People do the same thing. They'll sit there and present you 
>>>>> with a solution that is ridiculous because of some insider 
>>>>> information you have. I guess I'll tell these kinds of 
>>>>> people something like, "Look bud, I told you that won't 
>>>>> help. Are you listening to me?"
>>>>
>>>>
>>>> i guess its all YOU then. YOU dont want to have clarity but 
>>>> live in a grey zone of obsurity ambiguity.
>>>>
>>>> if you feel that is your safe zone, then fine. but to expect 
>>>> germination on air and mist is impossible.. unless you are an 
>>>> orchid.... or fungus.
>>>>
>>>
>>> Oh no I wasn't talking about your solution being untenable. I 
>>> guess if I were Brennus I would have worked that in and 
>>> polished my charming piece of shit to hand you.
>>>
>>
>> he is not in this thread tho. anyway, its fine.
>>
> He started this stuff where he says nonsense in the third person 
> directed @ me. I don't like when people around me act like looney 
> tunes.

you should talk to him in 3rd person as well, its a tactic you 
could try.

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


#505422

From% <pursent100@gmail.com>
Date2026-02-16 14:08 -0700
Message-ID<fZqcnUANsPqhGg70nZ2dnZfqn_adnZ2d@giganews.com>
In reply to#505414
phoenix wrote:
> jojo wrote:
>> phoenix wrote:
>>> jojo wrote:
>>>> phoenix wrote:
>>>>> phoenix wrote:
>>>>>> phoenix wrote:
>>>>>>> jojo wrote:
>>>>>>>> phoenix wrote:
>>>>>>>>> LLM read jojo over and over as if Hitlar's veganism is a 
>>>>>>>>> talking point. AI don't know obscure.
>>>>>>>>
>>>>>>>> you have to know what you are asking first. you dont know, 
>>>>>>>> unfortunately.
>>>>>>>>
>>>>>>>> but you can ask the model to refine your question, ask it to 
>>>>>>>> give you 5 questions based on your framing.
>>>>>>>>
>>>>>>> Dude I use AI just as much as you. Twice I've been working on 
>>>>>>> problems with AI when they get fixed into a rut saying it's this 
>>>>>>> or that, your obscure solution is fundamentally powerless. Maybe 
>>>>>>> I need to be meaner with my AI and fully correct them every time 
>>>>>>> they make a mistake.
>>>>>>
>>>>>> And it's important too. The AI is suggesting a two-dimensional 
>>>>>> array where I have two one-dimensional arrays. This is a 
>>>>>> fundamental slowdown that I refuse to tolerate rather getting my 
>>>>>> way to work by keeping one of the lists up to date on some stuff.
>>>>>
>>>>> In jojo language this is N^2 data points where mine has 2*N. I 
>>>>> don't particularly like N^2 complexity in anything because I deal 
>>>>> with big data.
>>>>
>>>> you have to treat it like a tool, like a scissor. you dont put the 
>>>> scissor on the paper and expect things to get cut.
>>>>
>>>> that is what the whole domain of prompt engineering deals with.
>>>>
>>> It's possible. I want to tell the AI I care about speed more than 
>>> anything, even correctness.
>>
>> ask it, am i making sense and present me with options about what you 
>> think i mean.
>>
> People do the same thing. They'll sit there and present you with a 
> solution that is ridiculous because of some insider information you 
> have. I guess I'll tell these kinds of people something like, "Look bud, 
> I told you that won't help. Are you listening to me?"

nobody listens to a pair of eagle asses in a power dive

[toc] | [prev] | [standalone]


Back to top | Article view | alt.slack


csiph-web