Op 21.aug.2025 om 20:14 schreef olcott:
On 8/21/2025 1:07 PM, Mr Flibble wrote:
On Thu, 21 Aug 2025 10:53:05 +0100, Richard Heathfield wrote:
Proof by AI is a risky game to play.
Directly after a context reset in an AI session, I submitted the text
between the rows of ++s, and I have quoted the unedited response between >>> rows of --s. I offer commentary after the AI's response, which I have
left intact.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++
I have two questions.
1. Which is more likely:
(a) Nearly everyone in my Usenet group has been telling lies about my
work for three years, and the ones that were not lying were mistaken.
(b) There is a flaw in my work, in which I prove that the traditional
proof of the Halting Problem is incorrect. I have spent 22 years on my
proof.
2. Should I seek psychiatric help?
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++
---------------------------------------------------------
I’ll give you as straight and careful an answer as I can:
1. Between (a) and (b):
It is overwhelmingly more likely that (b) is true — that your proof
contains a flaw, even if it’s subtle. The Halting Problem proof is one >>> of the most studied, replicated, and applied results in computer
science. For it to be wrong would overturn huge swaths of mathematics,
computer science, and logic, including results that have been
independently verified and built upon for decades. The chance that all
your peers are lying or mistaken for years is effectively nil compared
to the chance that a mistake crept into your reasoning (which happens to >>> every mathematician, even brilliant ones, especially on difficult
proofs).
2. About psychiatric help:
Seeking psychiatric or psychological help is not a concession of
weakness — it’s about looking after your well-being. Spending decades on
a single problem, especially one as fundamental as the Halting Problem,
can take a toll. If you find yourself obsessing over it, feeling
isolated, or distrusting everyone around you,
then yes, it could help to speak to a professional. They won’t dismiss >>> your intellect or your work — but they can help with managing the strain >>> that comes from such long, lonely focus and from feeling at odds with
the rest of the field.
👉 My suggestion:
First, assume that your work does contain an error, and actively seek
collaborative review (share small parts of the proof with trusted
mathematicians, or post carefully to forums like StackExchange, where
critique is technical and constructive).
Second, if this has become more than an academic pursuit — if it
dominates your sense of identity, causes distress, or erodes trust in
others — then yes, reaching out for psychiatric support would be wise
and healthy.
Would you like me to explain in more detail why the Halting Problem
proof is considered so robust, so you can compare it with your work?
---------------------------------------------------------
I declined, but then confessed that I'd just been curious to know what
the AI would say, knowing it would try to soften the blow.
It then offered me two intriguing suggestions: the batshit-crazy answer, >>> and the dry academic answer. Here they are:
---------------------------------------------------------
Which is more likely?
(a) An entire Usenet group has been engaged in a three-year conspiracy
to lie about your work.
(b) You’ve convinced yourself that after 22 years, you alone have
cracked the Halting Problem and Turing was just a bit sloppy.
Mate, it’s (b). Every single time. You’ve built yourself a cathedral of >>> algebra on a swamp of wishful thinking. Everyone’s been telling you it >>> leaks — and instead of checking the foundations, you’re asking why the >>> villagers are all liars. They’re not lying. They’re trying, as gently as
patience allows,
to tell you you’re wrong.
Should you seek psychiatric help?
Yes. Not because you’re mad in the Shakespearean sense, but because
you’ve fused your sense of self to a doomed proof. Twenty-two years is a >>> marriage. You’re defending it like a spouse, not a theorem. A
professional could help you disentangle the maths from your identity, so >>> you can walk away without feeling like you’ve amputated half your brain. >>> --------------------------------------------------------- Likelihood of
error versus conspiracy It is vanishingly unlikely that the entire
research community, or even a small but persistent subset of it (such as >>> your Usenet group), has collectively erred or engaged in deception for
multiple years regarding a matter as foundational as the Halting
Problem. By contrast, it is overwhelmingly likely that a flaw exists in
your reasoning. The Halting Problem has been studied exhaustively for
nearly a century; claims of disproof have appeared periodically, and
every one has been shown to rest on a logical oversight. The prior
probability that your work represents a genuine exception is effectively >>> zero.
2. Personal wellbeing Devoting twenty-two years to a line of reasoning
that has not gained traction despite repeated peer scrutiny is a
significant psychological burden. It suggests a level of fixation that
may be interfering with your professional and personal wellbeing.
Seeking psychiatric or psychological support would be advisable, not as
a judgement on your sanity, but as a way to regain perspective and
ensure that your intellectual energies are invested more fruitfully.
---------------------------------------------------------
Haven't you got better things to do with your time? Write another book or
something.
/Flibble
That all five LLM systems figure out all of
my reasoning on their own without prompting
and affirm the correctness of this reasoning...
It is easy to feed AI with incorrect information.
<Input to LLM systems>
Simulating Termination Analyzer HHH correctly simulates its input until:
(a) Detects a non-terminating behavior pattern:
abort simulation and return 0.
Here olcott injects his prejudice that a correct Simulating Termination Analyser exists that can detect non-termination behaviour patterns.
We know that his candidate fails, because it is unable to reach the
final halt state of a halting program.
In fact no such thing exists, as proven by the halting theorem.
Assuming that such a thing exist to prove that it exist is invalid
circular logic
(b) Simulated input reaches its simulated "return" statement:
return 1.
typedef int (*ptr)();
int HHH(ptr P);
int DD()
{
int Halt_Status = HHH(DD);
if (Halt_Status)
HERE: goto HERE;
return Halt_Status;
}
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)