XPost: comp.theory, sci.logic, sci.math
From: dbush.mobile@gmail.com
On 10/22/2025 9:47 AM, olcott wrote:
> On 10/22/2025 8:00 AM, dbush wrote:
>> On 10/22/2025 8:48 AM, olcott wrote:
>>> On 10/22/2025 7:25 AM, dbush wrote:
>>>> On 10/22/2025 7:56 AM, olcott wrote:
>>>>> On 10/20/2025 10:20 PM, Kaz Kylheku wrote:
>>>>>> On 2025-10-21, olcott wrote:
>>>>>>> On 10/20/2025 9:11 PM, Kaz Kylheku wrote:
>>>>>>>> On 2025-10-21, olcott wrote:
>>>>>>>>> On 10/20/2025 8:27 PM, Kaz Kylheku wrote:
>>>>>>>>>> On 2025-10-20, olcott wrote:
>>>>>>>>>>> On 10/20/2025 4:03 PM, Kaz Kylheku wrote:
>>>>>>>>>>>> On 2025-10-20, olcott wrote:
>>>>>>>>>>>>> On 10/20/2025 1:29 PM, Kaz Kylheku wrote:
>>>>>>>>>>>>>> On 2025-10-20, olcott wrote:
>>>>>>>>>>>>>>> On 10/19/2025 2:39 PM, Kaz Kylheku wrote:
>>>>>>>>>>>>>>>> On 2025-10-19, dart200
>>>>>>>>>>>>>>>> wrote:
>>>>>>>>>>>>>>>>> i don't get y polcott keep hanging onto ai for dear
>>>>>>>>>>>>>>>>> life. anyone with
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> Throngs of dumb boomers are falling for AI generated
>>>>>>>>>>>>>>>> videos, believing
>>>>>>>>>>>>>>>> them to be real. This is much the same thing.
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> AI is just another thing Olcott has no understanding of.
>>>>>>>>>>>>>>>> He's not
>>>>>>>>>>>>>>>> researched the fundamentals of what it means to train a
>>>>>>>>>>>>>>>> language
>>>>>>>>>>>>>>>> network, and how it is ultimately just token prediction.
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>> It excels at generating good syntax. The reason for that
>>>>>>>>>>>>>>>> is that the
>>>>>>>>>>>>>>>> vast amount of training data exhibits good syntax.
>>>>>>>>>>>>>>>> (Where it has bad
>>>>>>>>>>>>>>>> syntax, it is idiosyncratic; whereas good syntax is
>>>>>>>>>>>>>>>> broadly shared.)
>>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> I provide a basis to it and it does perform valid
>>>>>>>>>>>>>>> semantic logical entailment on this basis and shows
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> But you're incapable of recognizing valid entailment from
>>>>>>>>>>>>>> invalid.
>>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> Any freaking idiot can spew out baseless rhetoric
>>>>>>>>>>>>> such as this. I could do the same sort of thing
>>>>>>>>>>>>> and say you are wrong and stupidly wrong.
>>>>>>>>>>>>
>>>>>>>>>>>> But you don't?
>>>>>>>>>>>>
>>>>>>>>>>>>> It is a whole other ballgame when one attempts
>>>>>>>>>>>>> to point out actual errors that are not anchored
>>>>>>>>>>>>> in one's own lack of comprehension.
>>>>>>>>>>>>
>>>>>>>>>>>> You don't comprehend the pointing-out.
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> You need to have a sound reasoning basis to prove
>>>>>>>>>>> that an error is an actual error.
>>>>>>>>>>
>>>>>>>>>> No; /YOU/ need to have sound reasonings to prove /YOUR/
>>>>>>>>>> extraordinary claims. The burden is on you.
>>>>>>>>>>
>>>>>>>>>> We already have the solid reasoning which says things are
>>>>>>>>>> other than as
>>>>>>>>>> you say, and you don't have the faintest idea how to put a
>>>>>>>>>> dent in it.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> In other words you assume that I must be wrong
>>>>>>>>> entirely on the basis that what I say does not
>>>>>>>>> conform to conventional wisdom.
>>>>>>>>
>>>>>>>> Yes; you are wrong entirely on the basis that what you say does not
>>>>>>>> follow a valid mode of inference for refuting an argument.
>>>>>>>>
>>>>>>>> If you are trying to refute something which is not only a widely
>>>>>>>> accepted result, but whose reasoning anyone can follow to see it
>>>>>>>> for themselves, you are automatically assumed wrong.
>>>>>>>>
>>>>>>>> The established result is presumed correct, pending your
>>>>>>>> presentation of a convincing argument.
>>>>>>>>
>>>>>>>> That's not just wanton arbitrariness: your claims are being
>>>>>>>> directly refuted by elements of the established result which
>>>>>>>> we can refer to.
>>>>>>>>
>>>>>>>> I cannot identify any flaw in the halting theorem. It's not simply
>>>>>>>> that I believe it because of the Big Names attached to it.
>>>>>>>>
>>>>>>>
>>>>>>> And when I identify a flaw yo simply ignore
>>>>>>> whatever I say.
>>>>>>
>>>>>> Nope; all the ways you say claim you've identified a flaw have been
>>>>>> dissected by multiple poeple to a much greater detail than they
>>>>>> deserve.
>>>>>>
>>>>>> It is disingenuous to say that you've simply had your details
>>>>>> ignored.
>>>>>>
>>>>>
>>>>> Turing machines in general can only compute mappings
>>>>> from their inputs. The halting problem requires computing
>>>>> mappings that in some cases are not provided in the
>>>>> inputs therefore the halting problem is wrong.
>>>>
>>>> False:
>>>>
>>>> (,Y) maps to 1 if and only if X(Y) halts when executed directly
>>>> (,Y) maps to 0 if and only if X(Y) does not halt when executed
>>>> directly
>>>>
>>>
>>> Yes that it the exact error that I have been
>>> referring to.
>>
>> That is not an error. That is simply a mapping that you have admitted
>> exists.
>>
>>>
>>> In the case of HHH(DD) the above requires HHH to
>>> report on the behavior of its caller
>>
>> False. It requires HHH to report on the behavior of the machine
>> described by its input.
>>
>
> That includes that DD calls HHH(DD) in recursive
> simulation.
Which therefore includes the fact that HHH(DD) will return 0 and that DD
will subsequently halt.
--- SoupGate-Win32 v1.05
* Origin: you cannot sedate... all the things you hate (1:229/2)
|
