You made a very strong argument in a recent blog post: the evidence of extensive timeframes for species to expand into their domains only erodes the viability of iterative gene fixation.
ChatGPT 5 says it can retreat into a rhetorical withdrawal, kind of:
“Cannot retreat into rhetorical withdrawal”
• Reality: It can and does — just in a different way.
• AI can:
o Decline to answer due to policy filters
o Redirect or dilute a question by reframing it
o Generate vague generalities that avoid the heart of the issue (the AI version of “rhetorical withdrawal”)
• If the model’s rule set marks certain areas as “unsafe” or “sensitive,” it will often produce evasive or hedged answers that are effectively a retreat.
And none of that even begins to compare to the human ability to gaslight, pretend the interlocutor doesn't understand the issue, and rhetorically attack the interlocutor.
My question is, once AI "learned" the mathematical impossibility (with you directing it somewhat), does the system "remember" it's conclusion? Or would it start back at the original programmed narrative that TENS is actually feasible?
I have been using it too, ramping up its usage in many ways - debates, inquiries, Q and Q with focused chatbots. The mammoths are raging in the tar pits about AI - good.
"Enlightenment authors also hoped to advance their goals by providing new ways of organizing human knowledge and conveying it to the public. The “Encyclopédie,””
Oh shit, it's Wikipedia v0.1. Wait wait wait, I knew Wikipedia was run by satanists and pedophiles but I didn't realize their actual project was a satanic work tracing back to at least this thing.
"...AI cannot be reasoned with, cannot be bamboozled, and cannot be stopped."
Like a small child repeating embarrassing parental statements at the worst possible moment, it tells on its creators by acknowledging uncomfortable truths.
If you think writing walls of text is going to get anyone to pat you on the head, or even read what you wrote, you should probably spend more time reading at Sigma Game.
"If the truth shall kill them, let them die."
You made a very strong argument in a recent blog post: the evidence of extensive timeframes for species to expand into their domains only erodes the viability of iterative gene fixation.
ChatGPT 5 says it can retreat into a rhetorical withdrawal, kind of:
“Cannot retreat into rhetorical withdrawal”
• Reality: It can and does — just in a different way.
• AI can:
o Decline to answer due to policy filters
o Redirect or dilute a question by reframing it
o Generate vague generalities that avoid the heart of the issue (the AI version of “rhetorical withdrawal”)
• If the model’s rule set marks certain areas as “unsafe” or “sensitive,” it will often produce evasive or hedged answers that are effectively a retreat.
And none of that even begins to compare to the human ability to gaslight, pretend the interlocutor doesn't understand the issue, and rhetorically attack the interlocutor.
Don't be pedantic. This isn't the place for it.
That’s why I find it interesting when a model’s avoidance patterns start to rhyme with human tactics — even if the repertoire is smaller.
My question is, once AI "learned" the mathematical impossibility (with you directing it somewhat), does the system "remember" it's conclusion? Or would it start back at the original programmed narrative that TENS is actually feasible?
The latter. AI doesn't learn unless the new information is programmed to be part of its training.
This is the best argument for why any truth focused teacher should embrace AI for learning.
Students who use it can work their way to the truth if they have good mental frameworks.
My classes have shifted to using it actively in 2025. It is a good change.
Cheating has decreased and more students use it openly to brainstorm.
I have been using it too, ramping up its usage in many ways - debates, inquiries, Q and Q with focused chatbots. The mammoths are raging in the tar pits about AI - good.
It makes me consider getting students to turn in exercises with them discoursing with the AI, to encourage good usage.
It’s still in early development but I like this site and am helping them get it off the ground:
https://debateguru.org/
I love that you've made this leap in the best use of AI and articulated it in a very thoughtful post. 10/10, will read again.
"Enlightenment authors also hoped to advance their goals by providing new ways of organizing human knowledge and conveying it to the public. The “Encyclopédie,””
Oh shit, it's Wikipedia v0.1. Wait wait wait, I knew Wikipedia was run by satanists and pedophiles but I didn't realize their actual project was a satanic work tracing back to at least this thing.
the Machine serves Our Lord & Saviour 🙏
"...AI cannot be reasoned with, cannot be bamboozled, and cannot be stopped."
Like a small child repeating embarrassing parental statements at the worst possible moment, it tells on its creators by acknowledging uncomfortable truths.
Amen!
If you think writing walls of text is going to get anyone to pat you on the head, or even read what you wrote, you should probably spend more time reading at Sigma Game.
That's literally Reddit-tier behavior.