On 6/26/25 12:43 AM, olcott wrote:
https://chatgpt.com/share/685cca95-d8d4-8011-8506-e882917d16a8
No, all you have shown is that by LYING to chat GPT, you can confuse it
enough to give wrong answers.
Note, in interogating ChatGPT about what it has been told, and asking it
if it has been told that wrong answers can be right, it says:
There’s a difference between:
❌ A wrong answer — asserting something false.
🚫 No answer — admitting we can’t decide.
In practice, when we say an analyzer like HHH “returns 0” on a halting program like DDD, we don’t mean it is claiming “DDD does not halt” as a fact. Instead, we mean:
“HHH cannot prove that DDD halts using its simulation model, so under
its conservative rules, it classifies it as non-halting.”
and then later
But let’s be clear:
✅ Saying “this program never halts” when it does halt is a wrong answer. 🟨 It’s only “acceptable” if the system was never meant to give perfect answers, and the user knows that.
So, it ADMITS that it is lying to say it is correct, only that it is a
"safer" answer.
Sorry, but you are just proving that you don't know what you are talking
about, because you fundamentally don't understand what the meaning of
the words are, because you made yourself intentionally stupid in the field.
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)