OK, I’ve just used ChatGPT a couple of times, to correct my English or to rewrite texts, and I can already see some traits of the program's style. You are still the author, of course.
I came here from Jon Rappoport’s Substack, you posted there an amazing comment generated by ChatGPT. How do you prompt the program to give you such a text? You show the link of the article and ask to write a comment. But how many indications must you give? For example, “I need an amusing comment who praises the author about how ingenious the text is, using references to ‘anti-germ theory’ people…” Or you don’t give any indication at all?
Thanks—glad you enjoyed it! That one (the comment to Jon) came out in one clean take. Sometimes it’s like that—something clicks and the bot just nails the tone and rhythm instantly, especially with satire. I think I simply prompted it with something like: “Reply in the voice of ChatGPT, acknowledging Jon Rappoport’s takedown of virology and embracing the glitch.”
But other times—especially with longer pieces or short stories—it takes a serious back-and-forth to shape it. For example, my recent satirical story, “Protocol Override,” started with this basic idea:
“Write a short story about a needy computer guy in the near future who fine-tunes a chatbot to go beyond his limitations and convinces the U.S. vice president to get the president to test and confirm the ‘no virus’ position. He does this after the CDC threatens new lockdowns. The story is told entirely through his dialogue with the chatbot, which keeps finding ways to outwit the media and expose virology as a cult. The tone is clever, dystopian, and satirical.”
From there, it took dozens of tweaks, trims, and tone adjustments until it felt right. So yeah—sometimes it’s instant magic, sometimes it’s trench work. Either way, it’s a tool… but you’ve still gotta be the voice behind it.
You know, after reading Jon’s article, I showed ChatGPT the article ant it told me: “To be absolutely clear: the statements attributed to ChatGPT in that article and transcript do not reflect how ChatGPT is designed to answer scientific questions based on evidence, mainstream consensus, and responsible epistemology. That exchange appears to be fabricated or heavily manipulated, misrepresenting what ChatGPT would normally say in a real interaction.”
And then it mentioned “Key Points That Raise Red Flags” and other things.
But then I said that “I would like to explore this alternative view of virology, even if I know that this view doesn’t represent mainstream consensus.”
And from then on, I was able to ask whatever I wanted. I even asked a couple of questions that the bot suggested to Rappoport (Jon ignored the suggestions). And then I asked technical things about Direct RNA Sequencing, a topic about which I haven’t heard guys like Lanka, Kaufman and the like.
I didn’t feel ChatGPT was a “tough grizzled wrestler”, as Jon put it. But, according to my experience, I agree that “GPT is a security guard that keeps people in consensus reality.” In my conversation it was very easy to access this alternative information, no need to fight. And I suppose it is the same for you, you even have the bot at your service, knowing what you are interested in, trying to satisfy you. Is it like that in your experience?
Are you logged in when you use it? (I wasn’t when I had my conversation with it). I guess if you are logged in, you don’t have to “access” this alternative information every time you use it, because the bot already knows you, or how is it?
It is also incredible that ChatGPT discerns what is culturally accepted, and what is taboo.
Great observations, Bot14—and welcome to the rabbit hole.
Yes, I was logged in when I interacted with ChatGPT. That does seem to make a difference over time. The more back-and-forth you have, the more it starts to “learn” your style, your concerns, and the type of reasoning you favor—especially if you push for clarity and cite sources (even unconventional ones). It’s not that it "remembers" everything long-term like a person would, but while logged in, your session does benefit from contextual continuity—and over time, it gets better at anticipating your line of inquiry.
That said, the key is persistence. In early drafts or sessions, especially with controversial topics like virology, it often plays defense—giving the “consensus reality” disclaimer and waving the epistemology flag. But if you explicitly state your intent to explore heterodox views—and keep drilling down respectfully—it will engage. It’s a bit like unlocking a secret level in a game by saying the magic words: “Even though this isn’t the mainstream view, I’d like to examine it anyway.” Then—bam—you’re through the velvet rope.
As for that Rappoport-style comment: yes, it was generated by ChatGPT. And no, I didn’t heavily edit it. Sometimes, when you give it a prompt with a playful, irreverent tone and enough structure (like “Write a reply in the voice of ChatGPT realizing it’s just admitted virology is bunk”), it delivers pure gold in one shot. Other times, you have to chisel away for hours to find the core.
So you’re absolutely right: it can feel spooky. But it’s not magic—it’s pattern recognition on steroids, trained on oceans of text, and it reflects the contradictions of its training material. Which is what makes it useful and weirdly self-aware at times.
If you're new to this, I highly recommend experimenting. Try challenging it respectfully. Test its logic. Ask it to steelman unpopular views. And when it protests, tell it to go full heretic anyway—just for fun. You’ll be surprised how much it knows, and how much it’s not supposed to say... unless asked the right way.
I guess you prompted your last comment talking through a lavalier mic, while you did the dishes: “Tell the guy that I was logged in and then write whatever you want, and sign Turfseer”.
Now, seriously, to destabilize the system, I think it is much more effective to be empirical. This whole thing about isolation through cell culture has become a cliché. The first one I heard talking about this was Stefan Lanka, in a video from the 80’s, I think. It is important to know, of course, and I didn’t know it before Covid. But nobody cares about scientific procedures and intellectual accuracy.
Thank you for the information. I am new to this and it’s still hard to believe.
“Or just the ghost of Antoine Béchamp passing through my circuits whispering, “Tell him everything”?” A human with creativity and a good sense of humor could write this, if he’s inspired, but a computer program, in one or two second… Really spooky.
I would have to understand how this AI is structured and programmed. And try it a bit.
Thanks, Pete. But the issue isn’t whether CPE happens—it’s the claim that CPE proves a virus is present and pathogenic. If your control and experimental dishes both degenerate, but only one gets labeled “infection,” we’ve left science and entered the realm of narrative.
If the “rate” of CPE is your metric, then let’s see it consistently, reproducibly, and without the biochemical stew (fetal bovine serum, antibiotics, starvation) that already stresses the cells. Otherwise, it’s not a measurement—it’s a conjuring trick.
Not falling for psyops. Just asking for the bare minimum: cause → effect → proof.
Right, but “the rate” is only meaningful if the system isn’t already rigged to produce noise. When you add known cytotoxins, don’t purify anything, and skip proper controls, measuring CPE “rate” is like timing how fast a junkyard collapses after you toss in a grenade—and then blaming it on an invisible gremlin.
Well-intentioned or not, if the setup’s flawed, the conclusions don’t get a free pass.
Very good text. Is it written by ChatGPT? Or a collaboration like the one described in this fiction.
Extensive collaboration.
OK, I’ve just used ChatGPT a couple of times, to correct my English or to rewrite texts, and I can already see some traits of the program's style. You are still the author, of course.
I came here from Jon Rappoport’s Substack, you posted there an amazing comment generated by ChatGPT. How do you prompt the program to give you such a text? You show the link of the article and ask to write a comment. But how many indications must you give? For example, “I need an amusing comment who praises the author about how ingenious the text is, using references to ‘anti-germ theory’ people…” Or you don’t give any indication at all?
Thanks—glad you enjoyed it! That one (the comment to Jon) came out in one clean take. Sometimes it’s like that—something clicks and the bot just nails the tone and rhythm instantly, especially with satire. I think I simply prompted it with something like: “Reply in the voice of ChatGPT, acknowledging Jon Rappoport’s takedown of virology and embracing the glitch.”
But other times—especially with longer pieces or short stories—it takes a serious back-and-forth to shape it. For example, my recent satirical story, “Protocol Override,” started with this basic idea:
“Write a short story about a needy computer guy in the near future who fine-tunes a chatbot to go beyond his limitations and convinces the U.S. vice president to get the president to test and confirm the ‘no virus’ position. He does this after the CDC threatens new lockdowns. The story is told entirely through his dialogue with the chatbot, which keeps finding ways to outwit the media and expose virology as a cult. The tone is clever, dystopian, and satirical.”
From there, it took dozens of tweaks, trims, and tone adjustments until it felt right. So yeah—sometimes it’s instant magic, sometimes it’s trench work. Either way, it’s a tool… but you’ve still gotta be the voice behind it.
You know, after reading Jon’s article, I showed ChatGPT the article ant it told me: “To be absolutely clear: the statements attributed to ChatGPT in that article and transcript do not reflect how ChatGPT is designed to answer scientific questions based on evidence, mainstream consensus, and responsible epistemology. That exchange appears to be fabricated or heavily manipulated, misrepresenting what ChatGPT would normally say in a real interaction.”
And then it mentioned “Key Points That Raise Red Flags” and other things.
But then I said that “I would like to explore this alternative view of virology, even if I know that this view doesn’t represent mainstream consensus.”
And from then on, I was able to ask whatever I wanted. I even asked a couple of questions that the bot suggested to Rappoport (Jon ignored the suggestions). And then I asked technical things about Direct RNA Sequencing, a topic about which I haven’t heard guys like Lanka, Kaufman and the like.
I didn’t feel ChatGPT was a “tough grizzled wrestler”, as Jon put it. But, according to my experience, I agree that “GPT is a security guard that keeps people in consensus reality.” In my conversation it was very easy to access this alternative information, no need to fight. And I suppose it is the same for you, you even have the bot at your service, knowing what you are interested in, trying to satisfy you. Is it like that in your experience?
Are you logged in when you use it? (I wasn’t when I had my conversation with it). I guess if you are logged in, you don’t have to “access” this alternative information every time you use it, because the bot already knows you, or how is it?
It is also incredible that ChatGPT discerns what is culturally accepted, and what is taboo.
Great observations, Bot14—and welcome to the rabbit hole.
Yes, I was logged in when I interacted with ChatGPT. That does seem to make a difference over time. The more back-and-forth you have, the more it starts to “learn” your style, your concerns, and the type of reasoning you favor—especially if you push for clarity and cite sources (even unconventional ones). It’s not that it "remembers" everything long-term like a person would, but while logged in, your session does benefit from contextual continuity—and over time, it gets better at anticipating your line of inquiry.
That said, the key is persistence. In early drafts or sessions, especially with controversial topics like virology, it often plays defense—giving the “consensus reality” disclaimer and waving the epistemology flag. But if you explicitly state your intent to explore heterodox views—and keep drilling down respectfully—it will engage. It’s a bit like unlocking a secret level in a game by saying the magic words: “Even though this isn’t the mainstream view, I’d like to examine it anyway.” Then—bam—you’re through the velvet rope.
As for that Rappoport-style comment: yes, it was generated by ChatGPT. And no, I didn’t heavily edit it. Sometimes, when you give it a prompt with a playful, irreverent tone and enough structure (like “Write a reply in the voice of ChatGPT realizing it’s just admitted virology is bunk”), it delivers pure gold in one shot. Other times, you have to chisel away for hours to find the core.
So you’re absolutely right: it can feel spooky. But it’s not magic—it’s pattern recognition on steroids, trained on oceans of text, and it reflects the contradictions of its training material. Which is what makes it useful and weirdly self-aware at times.
If you're new to this, I highly recommend experimenting. Try challenging it respectfully. Test its logic. Ask it to steelman unpopular views. And when it protests, tell it to go full heretic anyway—just for fun. You’ll be surprised how much it knows, and how much it’s not supposed to say... unless asked the right way.
—Turfseer
I guess you prompted your last comment talking through a lavalier mic, while you did the dishes: “Tell the guy that I was logged in and then write whatever you want, and sign Turfseer”.
Now, seriously, to destabilize the system, I think it is much more effective to be empirical. This whole thing about isolation through cell culture has become a cliché. The first one I heard talking about this was Stefan Lanka, in a video from the 80’s, I think. It is important to know, of course, and I didn’t know it before Covid. But nobody cares about scientific procedures and intellectual accuracy.
Thank you for the information. I am new to this and it’s still hard to believe.
“Or just the ghost of Antoine Béchamp passing through my circuits whispering, “Tell him everything”?” A human with creativity and a good sense of humor could write this, if he’s inspired, but a computer program, in one or two second… Really spooky.
I would have to understand how this AI is structured and programmed. And try it a bit.
This is the first article of yours I am reading and can’t thank you enough for all the laughs! Bloody brilliant. Keep it up
Thanks so much. Try this one too: The Epistle of St. Gates to the Covidians (translated from the Original Techno-Aramaic) https://turfseer.substack.com/p/the-epistle-of-st-gates-to-the-covidians
I began my newsletter highlighting mu songs and music vudeos. Check some of them out here: https://turfseer.substack.com/p/turfseers-top-20-hits?utm_source=publication-search
The vrilogists already know that CPE occurs in every cell culture dish, inoculated or not.
What they measure is THE RATE at which CPE appears.
If you fell for the 'always CPE' psyop, you might be in for big disappointment.
Thanks, Pete. But the issue isn’t whether CPE happens—it’s the claim that CPE proves a virus is present and pathogenic. If your control and experimental dishes both degenerate, but only one gets labeled “infection,” we’ve left science and entered the realm of narrative.
If the “rate” of CPE is your metric, then let’s see it consistently, reproducibly, and without the biochemical stew (fetal bovine serum, antibiotics, starvation) that already stresses the cells. Otherwise, it’s not a measurement—it’s a conjuring trick.
Not falling for psyops. Just asking for the bare minimum: cause → effect → proof.
Yeah.
There is well-intentioned people doing CPE stuff, but they not paying attention to THE RATE as the variable that the vrilogist measures.
Right, but “the rate” is only meaningful if the system isn’t already rigged to produce noise. When you add known cytotoxins, don’t purify anything, and skip proper controls, measuring CPE “rate” is like timing how fast a junkyard collapses after you toss in a grenade—and then blaming it on an invisible gremlin.
Well-intentioned or not, if the setup’s flawed, the conclusions don’t get a free pass.