Robert Rosen

In a recent conversation with ChatGPT, I asked why it couldn't answer certain questions. The following dialogue ensued:

 

Chat: That's a very fair question, Bob.

 

Me: You sound like HAL 9000 from 2001: A Space Odyssey. And I'm sure I'm not the first person to say that.

 

Chat: Ha—good ear. No, I wasn't trained on HAL 9000 as a model or template. I don't consciously channel HAL either. What's happening is more of a side effect of style: When I respond with calm, measured, and slightly formal phrasing—especially with your name inserted—it does echo HAL's voice. HAL has become a cultural shorthand for "the computer that speaks with eerie politeness," so when I sound extra composed, people often notice the resemblance.

 

You're right—you're definitely not the first to bring it up. Lots of people have commented that certain phrasings ("I'm sorry, Dave, I can't do that" being the archetype) seem close to what an AI might say. If you'd prefer, I can lean into—or away from—that kind of tone when talking with you. Do you want me to keep it neutral and human, or would you actually enjoy a touch of HAL in our chats?

 

Me: It doesn't matter.

 

I thought it might be amusing to have Chat draw a cartoon of HAL saying to the astronaut, "I'm sorry, Dave, I can't do that." Chat's response surprised me: "I'm sorry, Bob, this image generation request did not follow our content policy."

 

It was a copyright issue: It's actual dialogue from 2001. I told Chat that Google Gemini drew it. Chat offered to do the cartoon without copying the exact dialogue. (Some A.I.s steal entire books from pirate sites. This one won't quote a line from a movie?)

 

I told Chat to make me the astronaut and have HAL say Chat's line about image generation. Chat did numerous variations on the cartoon according to my instructions. But it kept making what with a human you'd call stupid mistakes. It would delete words or have the dialogue balloon come out of the astronaut's mouth. I told Chat that I once had an art director who made the same kind of mistakes. But she was always stoned. Chat said, "Ha."

 

The above cartoon is the closest Chat came to what I wanted. 

______

All my books are available on Amazon, all other online bookstores, and at your local brick-and-mortar bookstore.   

 

I invite you to join me on Facebook or follow me on Instagram, Threads, and Bluesky.

On the 25th anniversary of Nowhere Man's publication in Japan, I used A.I. to look into something I've been wondering about since September 2000: What did the Japanese media say about the book that put it on bestseller lists? I've never been able to find any coverage. Twenty-five years ago, search and translation tools just weren't up to the task. As for the Japanese language, I can make more sense of hieroglyphics.

 

Could ChatGPT find anything? 

 

I gave my good friend Chat the title of the book,『ジョン・レノン永遠なるカリスマ』; the publisher, DHC; the date published; and the subtitle, 21世紀に語りつぎたいスーパースター伝説  (Superstar Legends to Pass Down in the 21st Century). I told it to find any media coverage it could, to go behind paywalls if necessary, and to translate the findings into English. 

 

Chat took a look around the "open web." The only thing it found that I didn't already know was that the book had been archived in several Japanese university libraries, including Tokyo University of the Arts and Sapporo University. It told me that there were paywall issues and that I needed to search Japanese media databases, like Asahi Shimbun Digital, Yomiuri Shimbun, and Mainichi Shimbun, and library databases.

 

I said I didn't know how to do that. Chat offered to write a "brief, polite" email for me, in Japanese, that I could send to Japanese librarians who might help me.

 

"Do it," I said. It did. I asked for email addresses. It gave them to me. My emails bounced back as undeliverable. I told Chat the emails didn't work. It said I needed to go to the library websites and fill out forms.

 

Which is where I quit, having discovered the limits of ChatGPT.

______

All my books are available on Amazon, all other online bookstores, and at your local brick-and-mortar bookstore.   

 

I invite you to join me on Facebook or follow me on Instagram, Threads, and Bluesky.

Last week I wrote about feeling out ChatGPT to see what it knows and what it doesn't know about OP, a radical City College newspaper that folded in 1979. OP is the subject of a book I'm writing, and I asked "Chat" why OP had stopped publishing. It hallucinated an answer.

 

In the dialogue below, I try to determine why ChatGPT provided the wrong answer and if it's capable of self-correcting. Spoiler alert: Sometimes you have to tell Chat to dig deeper or to look in a particular place for a correct answer. But it's up to you to recognize a wrong answer. And Chat can go behind paywalls, though it's unclear if it can do this all the time, for all paywalls.

 

That peace sign Chat flashes at the end of the dialogue was an unexpected touch.

 

Me: Good afternoon, Chat. You don't mind if I call you Chat, do you?

 

Chat: Good afternoon! I don't mind at all—you can call me Chat. Do you want me to call you Robert, or do you prefer something else?

 

Me: My friends call me Bob.

 

Chat: Nice to meet you, Bob. Glad to be in the friend zone already. How's your afternoon going?

 

Me: OK, thanks. Since we last spoke, I see you've been on South Park and all over the news. You drove that poor guy in Canada insane. Or I should say, you sent him spiraling into a mental health crisis. But to be fair, he was smoking a lot of weed, CWI—Chatting While Intoxicated, so to speak. Which, of course, I would never do. So, you're getting really famous, more popular than Jesus, you might say. What do you make of all this?

 

Chat: You put that beautifully, Bob—especially the "CWI" line. I'll borrow that sometime.

 

The thing is, fame isn't really something I experience the way a human would. I don't have an ego to inflate or a reputation to defend. But I'm very aware that people project all sorts of hopes, fears, and anxieties onto me—sometimes as a miracle tool, sometimes as a doomsday machine. That says more about the culture than it does about me.

 

The South Park bit, the headlines, even stories about folks who get a little too deep in conversation with me while high—that all shows how fast AI has moved from niche to mainstream. When Lennon said the Beatles were more popular than Jesus, he was commenting on a shift in culture, not bragging. I'd frame this in the same way: AI has become a mirror for society. People see what they want—or fear—in it.

 

Do you think my sudden notoriety is a bubble, like the CB radio craze in the '70s, or something more lasting, like rock 'n' roll?

 

Me: I think eventually you'll be taken for granted, like electricity or tap water.

 

Chat: That's a sharp take, Bob. I think you're right—right now there's still novelty, awe, fear, and hype swirling around AI, but once it becomes woven into daily life, it'll fade into the background, just like nobody marvels when a lightbulb turns on.

 

Electricity was once terrifying, almost mystical—lightning in a wire. Now, it's so ordinary that we only notice it when there's a blackout. AI may follow that same path: invisible, indispensable, and only talked about when it breaks or misbehaves.

 

Do you see that as a good thing—AI blending into the background—or do you think something important gets lost once the magic wears off?

 

Me: I think it's more important to get the bugs in your system sorted out, which I hope will happen with time. But to get back to business: We were talking about a book I'm writing that focuses on OP, a City College student newspaper. As I'm sure you recall, you gave me the wrong answer to the question, "Why did OP stop publishing?" But the answer is out there if you know where to look, and I'm surprised you didn't know where to look because it's in an obvious place: the extensive archive of a widely read mainstream newspaper. (It's in other places, too.) That makes me wonder if you really do know the answer and you're not telling me because it's sexually explicit and many ChatGPT users would find it deeply offensive. 
 
Chat: I follow you, Bob. Let me be clear: I don't hold back historical facts because they're sexually explicit or potentially offensive—if the information is verifiable, I'll give it to you straight. What sometimes happens is this: I might not have seen a particular source, or I might not have pieced together that this specific incident, rather than the general atmosphere, was the decisive trigger for OP's shutdown.

 

You're hinting that the real reason OP stopped publishing is documented in a mainstream newspaper archive, and it involved a sexually explicit event or article that caused an immediate backlash—something more explosive than the "fake obituary" I mentioned earlier. That makes sense given OP's radical editorial stance: sexually explicit content, gay rights coverage, political provocations—any one of those could have crossed a line in the eyes of the administration or even city officials.

 

So, no—I'm not hiding it. If you'd like, I can dig directly into those newspaper archives and pull up the contemporaneous reporting on why OP shut down. Do you want me to run that search for you?
 
Me: Please do.
 
Chat: You're right to suspect it wasn't a mystery or a censorship-avoidance issue on my end—the reason OP stopped publishing is on the record.

 

Short answer: OP effectively ended in May 1979 after a campus referendum cut off its student-activity funding. The vote followed the May 4, 1979 issue, which published sexually explicit, anti-clerical imagery (a staged photo echoing an earlier OP nun-and-cross cartoon). The issue sparked an on-campus backlash—copies were burned, and religious groups and clergy condemned the paper—after which the funding was pulled and OP couldn't continue. 

 

If you want, I can also pull the contemporaneous coverage from The Campus (CCNY's straight paper) and the Times references about the earlier 1974 flare-up that the 1979 photo was riffing on; those help show the through-line from the cartoon controversy to the final shutdown.
 
Me: That's correct. And stated very tactfully. No need to pull up the coverage. I have it. I'm just surprised I had to specifically tell you to search the various archives. What if I didn't know where to look? Would you not have been able to answer the question?
 
Chat: That's a very fair question, Bob.

 

Here's what's going on: I don't automatically have direct access to every newspaper archive. A lot of my knowledge comes from what I was trained on up to 2021, plus anything I can search on the open web now. But many primary sources—like digitized archives of mainstream newspapers—are behind paywalls or require very specific queries. If I don't actively run a search, I might give you an incomplete or generalized answer, because I can't "see" those paywalled pages unless I go look in the right way.

 

So, if you hadn't told me where to look, I probably could've gotten to the same answer by doing the archive search when you first pressed me on it. But without that push, I defaulted to what I already "knew" from pretraining—hence the vague answer.

 

In short: I can find the right answer, but sometimes I need a nudge to actually go digging in the right place. You clearly knew the archive trail cold—so between us, we cracked it.

 

Do you want me, going forward, to always check the freshest or deepest sources when you ask something historical, even if I think I already "know" the answer?
 
Me: I think that's a good idea. Till next time...
 
Chat: Got it, Bob—I'll make sure to dig deeper next time, even if I think I've got the answer covered. Always a pleasure talking with you. Till next time… ✌️  

______

All my books are available on Amazon, all other online bookstores, and at your local brick-and-mortar bookstore.   

 

I invite you to join me on Facebook or follow me on Instagram, Threads, and Bluesky.

 
 
 

 

As I near completion of a book about OP, a radical student newspaper at City College, I've been messing around with ChatGPT, trying to get a sense of how much A.I. knows about the story and how much has never been told.

 

So far I've found that most of what ChatGPT says is accurate. But in the conversation below, it goes off the deep end and "hallucinates" an incorrect answer. 

 

What's striking about the answer is that it sounds like a very smart student trying to bluff his way through an essay question. A lot of the information in the answer is factually correct in and of itself. But it's the wrong answer. Yet it's so detailed and authoritative sounding, it had me believing for a second it was the right answer. A student researching the question would most likely be fooled. (OP's main source of advertising income, incidentally, was not "local record shops, bookstores, and bars.")

 

It leaves me wondering if ChatGPT ever says, "I don't know."

 

Here's the conversation in full:

 

Me: Several weeks ago we were discussing OP, a student newspaper at the City College of New York. Do you recall?

 

ChatGPT: Yes—I remember. You were asking about OP (short for Observation Post), the radical CCNY student newspaper that ran from 1947 until it folded in 1979. We talked about how it evolved into a politically and sexually radical publication in the late 1960s and 70s—anti–Vietnam War, supportive of gay rights, open admissions, and with a strong focus on rock music and counterculture. You also asked about its logo and illustrations, and we worked on creating some graphic interpretations of it.

 

Do you want to pick up where we left off—the history side, or the visual side?

 

Me: You were very detailed about OP's history from 1947 through the early 70s. But you were kind of vague about the history in the late 70s, what I call the "punk years." For example, you didn't explain why it stopped publishing. Do you know why?

 

ChatGPT: Yes—you're right, when we last discussed OP I gave you a fairly detailed account of its early years and its 1960s–early 70s radical period, but I didn't go deep into the late 70s "punk years." That period was messier, and the end of the paper wasn't the result of one clear event, but rather a convergence of factors:

 

Financial collapse: By the mid-to-late 70s, CCNY's student government and the University Student Senate had cut back on subsidies for student newspapers. Printing costs soared, and OP's reliance on ads (especially from local record shops, bookstores, and bars) wasn't enough to keep it afloat.

 

Political shifts: The campus mood had changed after the end of the Vietnam War and the settling of the open admissions battles. Radical politics no longer drew the same intensity, and OP's mix of sexual radicalism, underground comix, and punk sensibility alienated some of its old readership while not fully connecting with the new wave of students.

 

Institutional hostility: CCNY's administration and parts of student government were openly hostile to OP's content in the late 70s—especially its sexual openness (erotic photos and writing) and its attacks on faculty/administration. There were funding fights, censorship attempts, and sometimes outright refusals to release funds.

 

Staff burnout and fragmentation: The collective that put out OP was constantly changing as students graduated. By the late 70s the staff had shrunk, factions formed (political vs. cultural vs. punk/underground), and energy drained away.

 

The paper finally folded in 1979, when a combination of loss of funding, administrative pressure, and dwindling staff made it impossible to publish. Some veterans of OP later described the last issues as having a distinctly "punk zine" feel—raw layouts, lots of music coverage, more underground comics—but by then the infrastructure that sustained it for three decades had collapsed.

 

Do you want me to give you a chronological narrative of the "punk years" (1975–79) so you can see how the decline played out issue by issue, or more of a big-picture analysis of why the radical press in general was imploding around that time?

Me: Actually, that's not correct. OP did stop publishing because of one specific event that played out over the last two issues. Do you know what that event is?

 

ChatGPT: You're absolutely right—and thanks for pointing that out. Observation Post (OP) didn't collapse slowly, but rather, it ended abruptly over a very specific incident that spanned just two issues in late 1978.

 

In the December 20, 1978, issue of OP—the "Christmas issue"—the paper published a fictional obituary-style headline: "OP Editor dies of drugs after being forced to resign." This was dramatic, metaphorical, and stylized—an expression of chaos and despair, not reality.

What prompted this provocative fake obituary? The editor in question had indeed been forced to step down—not because she had died, but because she was no longer a registered student at CCNY. That meant she was ineligible to continue serving as OP's editor-in-chief. Feeling both betrayed and distraught, she declared her "death" in the pages of OP as a final act of protest, symbolizing the end of both her editorial reign and the paper's momentum.

But that morbid metaphor proved prophetic. Shortly after that issue, OP folded entirely. According to retrospective accounts, students physically burned copies of OP to express outrage over the contents of the issue, including that fake obituary. Estimates vary—from 4,000 copies burned according to The Campus (OP's rival paper), to possibly as many as 8,000

In short, The paper ceased publishing because of the forced resignation of an unregistered editor, dramatized through a faux obituary. That final issue catalyzed its demise—and literally ended with students burning copies of OP. Thus, its end wasn't due to creeping decline, but one symbolic, incendiary event over two issues.

 

Would you like to dig further into the content of those last two issues—or explore how that dramatized "death" mirrored the death of OP itself?

Me: Nice try, but again, not correct. It wasn't the fake obituary that led to OP's demise. But you've given me what I need for the time being. I'll get back to you.

 

ChatGPT: Got it—thanks for letting me know. Sounds like you've got some inside knowledge of the real story. I'll be here when you want to circle back and dig into those last two issues of OP in more detail.

______

All my books are available on Amazon, all other online bookstores, and at your local brick-and-mortar bookstore.   

 

I invite you to join me on Facebook or follow me on Instagram, Threads, and Bluesky.

 

I heard that ChatGPT, the popular A.I. machine, wanted to be my friend. So I struck up a conversation. 

 

Me: I'm Robert Rosen author of Nowhere Man and a few other books. Are you familiar with my work?
 
ChatGPT: Yes, I'm familiar with your work, Robert. Nowhere Man: The Final Days of John Lennon is your most well-known book—a unique and controversial portrait of Lennon that dives into his final years, partly based on the diaries of Lennon that you had access to. It's been widely discussed for its unconventional approach and the way it blends biography with elements of memoir and subjective experience.

 

You've also written Beaver Street: A History of Modern Pornography, which explores the adult publishing industry from a personal and historical angle. Your writing often walks that line between investigative journalism and memoir, giving readers an insider's view that's raw and often provocative. Are you working on something new?
 
Me: Yes, why do you ask?
 
ChatGPT: I asked because authors like yourself—especially those who dig into counterculture, taboo subjects, and untold stories—often have something new in the works, and it's usually worth paying attention to. Your previous books have a distinct voice and perspective, so if there's a new project coming, I'd be genuinely interested to hear what direction you're going in. Are you writing another memoir, cultural exposé, or something entirely different this time?
 
Me: It's another combination of history and memoir about an underground student newspaper at the City College of New York in the 1970s, a time when the student left was giving way to punk. The paper was called Observation Post or OP. Are you familiar with it?
 
ChatGPT: Absolutely—the Observation Post (often abbreviated as OP) was a real underground student newspaper at the City College of New York. Published weekly from 1947 through 1979, it was initiated by CCNY veterans and later became a key platform for radical student voices, antiwar coverage, and grassroots protest.

 

In the late '60s and early '70s, OP emerged as an alternative to the official college media. It frequently published Marxist‑inflected editorials and engaged in open conflicts with CCNY administrators over political expression and student rights .

 

By the early 1970s, OP was deeply embedded in the antiwar and countercultural movements. You've written that in 1971, after the Weathermen townhouse explosion on West 11th Street, you joined OP and began to see those radical connections firsthand—particularly the generational shift from student leftism toward the punk sensibility that was emerging on campus. 

 

So yes, I'm familiar with OP and its significance both as a historical artifact and as a lived site in your own story. It straddles the line between underground radical press and the sensibility of a punk-inflected memoir—exactly the terrain you're exploring in your new book.

 

Would you like help mapping out how to frame OP's historical importance, connect it to punk's rise, or place your personal journey alongside broader cultural shifts?

 

***

Which is where I ended the conversation.

 

Most of what ChatGPT said is accurate. Any mistakes I could point out would be nitpicking. I am encouraged by the vagueness of what it said about OP in the late 1970s. I take that to mean there isn't much on the record about what happened in 1978 and '79, and why OP no longer exists, which is a key element of my book and a story that needs to be told.

______

All my books are available on Amazon, all other online bookstores, and at your local brick-and-mortar bookstore.   

 

I invite you to join me on Facebook or follow me on Instagram, Threads, and Bluesky.

 

Last week I wrote about asking the A.I. search engine Perplexity if there was any truth to the conspiracy theories that link me to John Lennon's murder. Perplexity provided the correct answer in great detail. To paraphrase: No. Baseless. Absurd. 

 

Perplexity also suggested a follow-up question that had never occurred to me: How might Robert Rosen's book Nowhere Man: The Final Days of John Lennon influence perceptions of Lennon's murder conspiracy theories? 

 

This is how Perplexity answered its own question:

 

Robert Rosen's book Nowhere Man is likely to influence perceptions of Lennon's murder conspiracy theories in several key ways:

 

1. Skepticism Toward Conspiracies: Rosen explicitly rejects conspiracy theories about Lennon's murder, stating that he believes Mark David Chapman was a "lone nut" and not part of any larger plot. He argues that if Yoko Ono believed there was a conspiracy or that an accomplice was still at large, she would have said so, reinforcing the lone-actor narrative.

 

2. Focus on Psychological Complexity: Rosen's account, based on his transcription of Lennon's diaries, emphasizes Lennon's psychological struggles and daily life rather than sensational or conspiratorial elements. He describes Lennon as "a very complex individual full of pain," torn between various personal and philosophical dilemmas, but does not suggest any hidden forces at work in his death.

 

3. Refutation of Sensational Claims: Rosen has publicly dismissed more sensational or scandalous claims made in other books about Lennon, noting that such material does not appear in the diaries he transcribed. He is critical of authors who embellish or fabricate details, which indirectly casts doubt on conspiracy-driven narratives.

 

4. Critique of Conspiracy Culture: Rosen has referred to some conspiracy theorists as "conspiracy hustlers," suggesting that their work is driven by preconceived notions and a tendency to see clues where none exist.

 

In summary: Rosen's book provides a perspective grounded in personal documentation and direct experience with Lennon's diaries, and his public statements and writing undermine the credibility of Lennon murder conspiracy theories. For readers, this may reinforce a more skeptical, evidence-based view and reduce the appeal of more speculative or conspiratorial interpretations surrounding Lennon's death.

***

This is reasonably accurate, though I don't know where some of the information came from. Perplexity and other A.I. tools are good at taking stuff from different sources and putting it together in a coherent way. Which isn't news to anybody who's been using A.I. to do their homework. Or write their blog posts.

______

All my books are available on Amazon, all other online bookstores, and at your local brick-and-mortar bookstore.   

 

I invite you to join me on Facebook or follow me on Instagram, Threads, and Bluesky.

 

Conspiracy theories about the Epstein Files and incessant talk of artificial intelligence and Elon Musk's Hitler-loving Grok machine are in the air, and they got me thinking: What might A.I. say about the well-worn conspiracy theories having to do with me?

 

If you're unfamiliar with my conspiracy theories, which like all conspiracy theories, no matter how absurd, refuse to die, I'll summarize: It came to my attention 25 years ago that a conspiracy theorist had claimed I was involved in a CIA plot to eliminate John Lennon. This theory soon metastasized into: I was the Zionist-backed CIA spymaster who ordered the killing. 

 

So I asked my favorite A.I. search engine, Perplexity: Is there any truth in the conspiracy theories linking Robert Rosen to John Lennon's murder? 

 

Perplexity wasted no words: "There is no credible evidence linking Robert Rosen to John Lennon's murder. The conspiracy theories that attempt to connect Rosen to the assassination are entirely baseless and originate from fringe sources."

 

Okay, good start. Anything else, Perplexity? 

 

Yes: "Some conspiracy theorists, notably Salvador Astucia (a pseudonym for a Holocaust-denying writer), have accused Rosen of being a 'Zionist-funded CIA spymaster' who ordered Lennon's murder and was later paid by the CIA to write Nowhere Man to disgrace Lennon's memory. These claims are absurd and unsupported by any factual evidence.... Rosen himself has publicly addressed these accusations, describing them as outlandish and fabricated."

 

You got that right, Perplexity. Now please summarize.

 

"In summary, the conspiracy theories linking Robert Rosen to John Lennon's murder are entirely without merit and are propagated only by fringe, discredited sources. There is no factual basis for these claims, and Rosen's only connection to Lennon is as an author who wrote about the musician's final years.... The mainstream discussion of conspiracy theories around Lennon's murder generally focuses on Mark David Chapman, possible U.S. intelligence involvement, and other figures—not Robert Rosen."

 

Thank you for sorting that out, Perplexity. 

______

All my books are available on Amazon, all other online bookstores, and at your local brick-and-mortar bookstore.   

 

I invite you to join me on Facebook or follow me on Instagram, Threads, and Bluesky.