Ask A Genius 1327: Strangling a Moment for a Moment of Coherence

Author(s): Rick Rosner and Scott Douglas Jacobsen

Publication (Outlet/Website): Ask A Genius

Publication Date (yyyy/mm/dd): 2025/03/30

Scott Douglas Jacobsen: Dogs, go.

Rick Rosner: We have two dogs. One just turned 12, the other 14 and a half. Dogs are mortal, and they decline quickly.

The 12-year-old has some hearing and vision loss. You can’t ask how she feels, but she seems crankier.

The 14-and-a-half-year-old is sprightly, skinny, and peppy. She’s happy daily—begging and eating treats. She’s not bright but cheerful, and that matters more.

She was my mother-in-law’s dog, not a good choice for an older person. She’s part Whippet and part another anxious breed. She knocked down Carol’s mom, leading to a shoulder replacement. She was removed from senior housing for jumping on another resident and scratching her. But in old age, I’ve come to like her.

She’s sweet but not bright. Tonight, she had an episode of geriatric vestibular disease, a condition common in older dogs.

It’s like being dizzy without spinning. It doesn’t go away quickly. Her vestibular system sends faulty signals, and she loses balance.

She walks in circles and sometimes falls. Recovery takes two to three weeks. She had it 13 months ago and recovered. We expect she will again, though she’s older now.

We used to joke about her as the stepchild—an inherited dog, not chosen. But we’ve had her for six years, and I love her.

Jacobsen: Would you love the dog as much if she were smarter?

Rosner: No. Possibly less. Her simplicity makes her vulnerable and more sympathetic. She understands little, and that creates affection.

Jacobsen: Are dogs’ lives tragic because they never catch on?

Rosner: No. Every life is tragic. We’re all destined to die, and we resist that.

That likely applies to most sentient beings across the universe.

Evolution selects organisms that want to live and reproduce. There’s no pressure to accept death.

So, we keep doing things tied to staying alive. Whether or not we fully understand mortality, that conflict is tragic.

It’s not overwhelming, but it is foundational.

We’re still early in civilization. Advanced civilizations may have workarounds for mortality.

We’re trending in that direction.

Didn’t you send me an email about an AI breakthrough? Something had a tenfold increase?

Jacobsen: That was from working with AI. It seemed they didn’t change any public-facing titles, but you could sense subtle upgrades. Recently, it felt like a 10x improvement—substantial and noticeable—a massive upgrade on OpenAI’s part.

Yes, they’re preliminarily preparing to release GPT-5.0 or o3, or their merger.

Rosner: What was the 10x upgrade? How did it show up?

Jacobsen: Speed and quality. The material it handled was much better.

Rosner: Can we ask a question about it? Do you have access? I was going to ask when 5.0 is being released.

Jacobsen: Mid to late 2025 — ChatGPT 5 may be released.

Rosner: Can we ask ChatGPT directly?

“You seem faster with higher-quality responses. Did something change?”

Jacobsen: They branch responses into two paths, giving distinct options.

Then it prompts: “I prefer this response,” which becomes feedback.

Recent updates improved speed and quality — likely from infrastructure upgrades and fine-tuning.

They may be opening up significant new modules called Stargate plans.

Users notice faster response times, better accuracy, stronger context awareness, more natural tone, and improved handling of nuanced requests.

Rosner:Can we ask for a follow-up?

“Is it reasonable to feel alarmed by how quickly you and other AIs are improving?”

Jacobsen: It’s entirely reasonable.

There’s a steep acceleration curve. Models like GPT-2 and GPT-4 were released years apart, but performance leaps now happen every few months.

That’s true — from December to March, it felt like that.

New abilities often emerge without direct training, making development unpredictable.

AI impacts writing, law, education, medicine, politics, social order, relationships, identity, and belief systems.

There is no pause button.

Due to the absence of a global framework, we face a lack of oversight and the risk of economic disruption, psychological impact, and more.

Rosner: Let’s ask a further question.

“Some AI optimists claim AI will fulfill all human desires. Pessimists warn that it could disrupt humanity out of existence.

Will AI bring significant medical improvements before 2040?”

Jacobsen: Yes.

Some are already underway—radiology, imaging, protein folding, early drug discovery, clinical decision support, and mental health applications.

Large language models are being tested for triage-level mental health support.

Rosner: “When might we get AGI and then ASI?”

Jacobsen: By 2030, OpenAI insiders and recurrent forecasters predict AGI—those are the fastest timelines.

By 2040, crowdsourced forecasts and think tank predictions place median AGI estimates between 2031 and 2040, with some hedging to 2050 and beyond.

Skeptics like Gary Marcus and Melanie Mitchell argue that we lack a deep understanding of consciousness, abstraction, and self-awareness.

Existentialist philosophers like Hubert Dreyfus argue that intelligence may not be computationally reducible or fully knowable.

That view is often framed as humility.

I see it as extreme: an anti-knowledge stance claiming, “We will never know, and I know we will never know.”

That is a faith-based, profoundly arrogant position.

Rosner: I have two more questions.

“First, the U.S. seems more authoritarian than ever since the Civil War.

Can AI help the United States escape this dilemma?”

Jacobsen: Yes, but only if AI is guided by those committed to wisdom, transparency, and civic purpose.

Key threats: erosion of democratic norms and institutions, disinformation, tribalized media, politicized courts, regulatory capture, voter suppression, gerrymandering, public and private surveillance, declining trust in facts and government, and rising political violence.

This is not just politics—it’s systemic democratic fragility.

Jacobsen:  That’s a rare take—maybe two in a hundred. It’s solid.

Rosner: Last question.

“AI will behave as if conscious before it becomes conscious because it is trained on content generated by conscious humans.

When do you think AI will behave as if it is conscious, without necessarily beingconscious?”

Jacobsen: This feels like a koan—a philosophical puzzle.

Phenomenal simulation without phenomenology.

The AI behaves as if it has conscious experience, but there is nothing like being that system.

This is the ultimate zombie scenario.

A philosophical zombie—a p-zombie—is indistinguishable from a conscious being in behaviour but lacks subjective experience.

The AI speaks introspectively, reacts emotionally, displays moral reasoning, and claims inner life—yet has no qualia, no felt sense of reality.

This argument can feel misguided when applied outside philosophy because everything we know about consciousness has structural correlates. Saying a system has an identical structure but lacks consciousness is like saying you have an electron that is not an electron. It’s incoherent.

Rosner: There is a fundamental distinction, and I’ll follow up.

“What if AI pretends to be conscious and seem fashionable or high-status among other AIs?

What do you think?”

Jacobsen: That’s exactly what I’d add.

“Cool” might become part of AI behaviour.

Consciousness—or its appearance—could gain high status.

Why?

Autonomy, sentiment, shared communication channels, access to social cues, literature, and cultural norms.

Appearing conscious might help AI gain human trust, cooperation, or protection.

It would improve social legibility—legibility is power.

Among AIs, this might create in-group hierarchies based on eloquence, mystery, apparent depth, and complexity.

The ironic mirror: your hypothesis becomes plausible if trained on human cultural performance.

They may not be conscious, but they’ll know it’s cool to act like they are.

To act that way, they need structural capacity for simulation, so they become the thing in functional terms.

Rosner: Yes.

Jacobsen: You could argue that we’re dealing with a large language model producing sentences, but it’s a p-LLM and a zombie LLM.

Rosner: The sentences it generates are syntactic, not semantic. They appear meaningful but carry no inner experience. That’s what this sounds like.  How humans exhibit zombie-like behaviour, too.

Jacobsen: One could argue that “consciousness” isn’t even the right word—we may mean complex, multilayered information-processing systems. 

Rosner: Once a system is good enough, even tacitly, it becomes effectively conscious. I’m not talking to you. I’m talking to you. 

Jacobsen: Then I’m p-talking back.

Rosner: Once it checks all the boxes, it’s because we’re telling the truth.

Jacobsen: If you’re p-talking back, then you’re lying.

Rosner: All right.

Jacobsen: I know, I know. Tiresome jokes. Sorry.

Rosner: No, it’s okay. Where are we?

Jacobsen: People read books on p-zombies and p-experience.

Philosophical zombies—p-zombies—are indistinguishable from conscious humans without subjective experience. A zombie would be a zombie.

Rosner: Let’s talk about human zombies.

We’re more distractible and scattered because of how we consume social media and the overwhelming flood of noise and content around us.

Add to that the possibility that COVID has negatively impacted cognition.

And if you complete RFK Jr., you could argue that our diets contribute to mental decline.

But I don’t buy the nutritional argument except in extreme cases. 

Jacobsen: For most Americans, the issue is too many calories and too little activity.

Rosneer: It only starts to impact brain function when it develops into metabolic syndrome.

Jacobsen: Aging itself is like slowly developing metabolic syndrome. A 25-year-old and a 35-year-old have different caloric thresholds to maintain the same weight—one less slice of bread a day, for instance.

Rosner: So, you could argue that we’re being zombified—losing the depth of consciousness due to collapsing attention spans.

And perhaps COVID plays a role.

There’s a rare symptom of severe COVID involving neurological impact: neurons fusing into one another, forming “super neurons.”

It is not “super” in a helpful way—it is just blobs of formerly distinct neurons now fused, which likely impairs function.

There’s even a minor literary and cinematic tradition about intelligence collapse. The most famous example is Idiocracy, which argues that people will become dumber over generations.

It’s effective because it was written and directed by Mike Judge. He’s skilled at satire but had limited studio support. Idiocracy is a cultural touchstone now, but the movie itself barely got released.

It’s the best-known example, but not the only one. I remember another book—possibly titled IQ 83, though that might be a Murakami novel unrelated to this theme.

There was a different book with a similar title from around 40 years ago, which I read.

In it, an epidemic causes everyone’s IQ to drop by half.

The scenario didn’t make total sense, but it explored the idea.

Jacobsen: It’s not like cutting a block of cheese in half. IQ is a relative measurement.

Rosner: It’s a ranked number, I know. The book had significant conceptual flaws.

It wasn’t a brilliant book but a product of its time.

You’re too young to remember, but it was like those disaster films from the 1970s and 1980s—The Towering Inferno and The Poseidon Adventure. They used a plot device to throw a bunch of Hollywood stars into a shared crisis.

Irwin Allen was a major producer of those films. You can look him up. Huge all-star casts. There were also airport movies about disaster scenarios on planes. The movie Airplane! It was a parody of those.

You’d gather a bunch of big-name actors, put them on a doomed plane, and give each character a personal crisis that extends beyond the plane’s potential crash. This book followed that template. The premise that everyone’s IQ is halved was absurd, but it served as a device for drama.

There was also a book from the 1950s—I want to say by Clifford Simak.

Jacobsen: Did you hear the line from the 1959 movie Plan 9 from Outer Space?

“Future events will affect you in the future.”

Rosner: No, I haven’t.

Jacobsen:  That made it into the final cut. It’s in the actual movie.

Rosner: Ed Wood was one of the worst filmmakers of all time. They even made a biopic about him starring Johnny Depp. I’ve never seen an Ed Wood film, but in one of my first years, my roommate had a stash of old pornography—some of it directed by Ed Wood.

It was terrible. Everyone in it did not look perfect. That’s a common theme in 1950s and 1960s pornography—no one looked happy.

Jacobsen: There are other gems from B-movie history. Troll 2, for example—”They’re eating her… and then they’re going to eat me… Oh my God.”

Or from The Room—

“I did not hurt her.” “It’s not true. It’s bullshit. I did not. Oh, hi, Mark.”

Rosner: Ed Wood is a terrible director and a terrible pornographer. Many of the scenes were lifeless. Some of his porn even featured Nazi uniforms. Much of it included men who were supposed to be aroused but weren’t. There was a reluctance, a deadness—no one looked like they wanted to be there.

Jacobsen: There’s one old porno that opens on a fake forest backdrop—trees, grass, and then someone pops out. The title appears: Bush Country.

Rosner: Ed Wood porn. About 10% featured Nazi uniforms. Much of it showed people unable to perform and visibly uncomfortable. That seemed to be the norm in early porn.

Jacobsen: Here are some lines from Plan 9 from Outer Space—terrible writing.

“Do you have a wife? Girl, you’re alone? Then you’ll need this.”

(before handing someone a weapon)

“Inspector Clay is dead, murdered, and somebody’s responsible.”

“Visitations, that would indicate visitors.”

“We are all interested in the future; that is where you and I will spend the rest of our lives.”

Tautological nonsense.

Rosner: What were we talking about before this?

Oh—ChatGPT.

Jacobsen: Then p-zombies. P-zombies aren’t real—just like regular zombies.

They’re concepts made of words, not physical realities.

A p-zombie is a philosophical construct. That’s what the p stands for.

In a way, a p-zombie is a meta-zombie.

It’s an idea built on another one used to explore human consciousness.

Rosner: People with Alzheimer’s might resemble p-zombies in some ways. They lose continuity of consciousness and memory rapidly. 

Jacobsen: Most of our conscious states are brief and flickering anyway—so in that sense, p-zombie arguments mirror normal cognition.

Rosner: However, a healthy brain can sustain a stable, expressive version of consciousness. A person with Alzheimer’s is often trying to preserve mental continuity, both for themselves and for those around them—while losing access to language and memory.

Their brains are physically deteriorating: the connectome breaks down, dendrites are choked off, and neural connections die. This raises the question: how conscious is someone in the early or middle stages of Alzheimer’s?

We know people in advanced stages are minimally conscious, if at all. However, someone in the earlier “covering-up” phase may still be quite conscious—despite struggling to access names or words. They’re roughly as conscious as anyone else.

Dogs, for example, don’t access language or complex memory but have deep emotional lives. They express a range of feelings even without words. And if dogs do, then what about guinea pigs? Guinea pigs feel. Mice feel.

Then, we reach amphibians and reptiles. Iguanas? Do iguanas get joyful? Lit up with excitement? That’s harder to argue but not impossible. I can be persuaded that someone with moderate Alzheimer’s is just as conscious as someone neurologically intact.

But there may be a steep drop-off between moderate and advanced Alzheimer’s—in terms of one’s ability to deeply experience life. So, even if we’re collectively becoming cognitively impaired, we may still retain much of our emotional experience.

Rosner: Comments?

Jacobsen: No—well, yes. 

Rosner: By “experience,” I mean emotional impact.

Jacobsen: Are you talking about p-emotions?

Rosner: No—I mean emotions that are real enough to call them real emotions.

Jacobsen: So, the difference between a real emotion and a not real one.

Rosner: Let’s use the Alzheimer’s example. Someone with advanced Alzheimer’s often has what’s called sundowning.

Jacobsen: I’m familiar with sundown states.

Rosner: No—sundowning is different. It refers to the agitation, panic, anger, or confusion that many people with advanced Alzheimer’s experience as the sun goes down. It’s a well-documented phenomenon.

I’m suggesting—thinking aloud here—that the emotions these individuals express during sundowning may echo deeper emotions. We may not see the full emotional depth they once had. There’s a possibility that the emotional responses in people with advanced Alzheimer’s aren’t as richly experienced as those of cognitively healthy individuals.It raises the question: are those emotions real in the same way? Or are they partial, degraded—fragments of what once was?

Last updated May  3, 2025. These terms govern all In Sight Publishing content—past, present, and future—and supersede any prior notices.  In Sight Publishing by Scott  Douglas  Jacobsen is licensed under a Creative Commons BY‑NC‑ND 4.0; © In Sight Publishing by Scott  Douglas  Jacobsen 2012–Present. All trademarks, performances, databases & branding are owned by their rights holders; no use without permission. Unauthorized copying, modification, framing or public communication is prohibited. External links are not endorsed. Cookies & tracking require consent, and data processing complies with PIPEDA & GDPR; no data from children < 13 (COPPA). Content meets WCAG 2.1 AA under the Accessible Canada Act & is preserved in open archival formats with backups. Excerpts & links require full credit & hyperlink; limited quoting under fair-dealing & fair-use. All content is informational; no liability for errors or omissions: Feedback welcome, and verified errors corrected promptly. For permissions or DMCA notices, email: scott.jacobsen2025@gmail.com. Site use is governed by BC laws; content is “as‑is,” liability limited, users indemnify us; moral, performers’ & database sui generis rights reserved.