Forums before death by AOL, social media and spammers... "We can't have nice things"
|    sci.logic    |    Logic -- math, philosophy & computationa    |    262,912 messages    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
|    Message 261,350 of 262,912    |
|    Chris M. Thomasson to All    |
|    Re: New formal foundation for correct re    |
|    26 Nov 25 21:58:44    |
      XPost: sci.math, comp.theory       From: chris.m.thomasson.1@gmail.com              On 11/25/2025 8:55 PM, dart200 wrote:       > On 11/25/25 8:23 PM, Python wrote:       >> Le 26/11/2025 à 05:22, dart200 a écrit :       >>> On 11/25/25 7:58 PM, Python wrote:       >>>> BTW you should think about what Ben Bacarisse once wrote:       >>>>       >>>> The set of all functions from ℕ to ℕ is uncountable (as large as the       >>>> real numbers), while the set of all finite programs is only       >>>> countable, so there are far more possible functions than there are       >>>> programs to compute them; this guarantees that most functions are       >>>> uncomputable and, more generally, that no finite formal system or       >>>> algorithmic procedure can cover “all” functions, all truths, or all       >>>> behaviors describable over the naturals—so whenever someone claims       >>>> to have a universal decider, a complete semantic engine, or a single       >>>> system that captures all “objects of thought,” they are implicitly       >>>> pretending that countably many programs can represent uncountably       >>>> many functions, which is mathematically impossible.       >>>>       >>>> The "halting problem" is actually only a way to confirm this with a       >>>> specific case.       >>>       >>> u don't need undecidable machines (that are actually hypothetical) to       >>> confirm uncomputable functions       >>>       >>> that is also something i stumbled upon in my musings       >>       >> So ?       >       > the "halting problem" is not the only way to confirm a specific case of       > an uncomputable function,              When shall your body on Earth halt?               and therefore it is not necessary for that       > reasoning       >       > in fact, the very first example of an uncomputable function was not an       > undecidable paradox like the halting paradox, read the first couple       > paragraphs of §8 from his 1936 paper       >              --- SoupGate-Win32 v1.05        * Origin: you cannot sedate... all the things you hate (1:229/2)    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
(c) 1994, bbs@darkrealms.ca