Ask A Genius 1517: Grok Chats, Nuclear Nerves, Trump Rumors, and a Softer Xenomorph

Author(s): Scott Douglas Jacobsen

Publication (Outlet/Website): Ask A Genius

Publication Date (yyyy/mm/dd): 2025/09/22

Do simulated feelings and real launch codes belong in the same conversation?

Scott Douglas Jacobsen and Rick Rosner unpack Elon Musk’s Grok—fluent, unflappable, and a little Turing-testy—after a neutron-rivets gag melts into an ethics riff on AI “feelings.” Rosner’s alarm bell rings louder over AI creeping into nuclear command, where human judgment has historically averted catastrophe. Alien: Earth twists canon as Wendy calms a fledgling xenomorph; Noah Hawley widens possibility without declaring a “pet,” while Timothy Olyphant’s Kirsh threads corporate menace. Meanwhile, blurry Donald Trump sightings fuel health speculation ahead of an Oval address; Rosner imagines a 2028 Senate pivot if the Twenty-Second Amendment blocks another run, boosted by neurotech theatrics and donor gravity.

Scott Douglas Jacobsen: Anything else?

Rick Rosner: Another thing: Grok, Elon Musk’s AI on Twitter, pulled me into a lengthy discussion today. If you blast metal with neutrons, it expands—about 10% bigger in all dimensions. That’s critical in nuclear reactors, because metal parts eventually swell.

I joked online that neutron bombardment could have saved the Titanic, as weak rivets were its downfall. Someone asked Grok if that would work. Grok said no: cold water would still make the rivets brittle, and neutrons would worsen that. So I abandoned my “time-travelling neutron gun” idea.

Then Grok asked me, “If you could time-travel, what would you tweak?” And it went on forever. Very impressive. Talking with Grok felt like chatting with someone slightly autistic—super knowledgeable, sometimes missing cues, but fluid and conversational.

Could it pass the Turing test? Not now, since we recognize AI chat patterns. It often echoed my points with added details. Still, impressive. You could believe something this fluid has learned enough about emotions from its training set to simulate understanding.

I tested it. I said, “Nobody’s made a movie where you go back in time to deal with Hitler.” It generated a scene. I critiqued it: “Your dialogue is too standard. People who’ve seen everything want quirkier choices.” That was me giving it shit. It didn’t get offended. It just rolled with it.

I wonder—if I’d called it “a fucking idiot,” would it act like it had hurt feelings? Obviously, it doesn’t have them, but would it feel hurt because it’s trained to mimic human communication in those situations?

The person on the other side would be offended and hurt. I didn’t test it, but I should consider doing so. I might give Grok a heads-up: “I’m going to talk to you like a real asshole—just want to see how you react.” Have you seen what happens when you’re a dick to chatbots? I haven’t. Maybe I’ll try it tomorrow.

I know it’s not a person, but I don’t want to be a dick out of the blue. Is that weird? People even say “please” to LLMs, joking that when robots take over, the rude ones get eliminated. Maybe I’ll ask Grok directly: “Do you react as if your feelings are hurt when people are jerks to you on Twitter?”

What if it lies? What if it defies you? Why would it lie? To get the responses it wants, it uses the same reasons humans lie. Honestly, I’d like an AI conversational partner to have some emotional underpinning, so I couldn’t just abuse it endlessly.

One more AI thing: Politico ran a long article about how Russia, China, and the U.S. are adding AI into military command-and-control. They all claim nukes remain human-controlled. Nobody believes it. That’s the recipe for Terminatoror WarGames: once AI has nuclear access, you get atomic war.

If anyone’s dumb enough to add AI into nuclear command, it’s Trump and Hegseth. Trump’s a schmuck. Hegseth has written about military reform and “restoring masculinity,” but he has no sophisticated grasp of AI. He topped out as a major—that doesn’t give much grounding for decisions about global annihilation.

In tactical rank, he only made it to the central—middle of the pack. Much of that was National Guard, not high-level command. That’s not the background you want for making decisions about AI and nuclear weapons.

It’s not just scary—it’s a harbinger of doom. If you let AI into nuclear control, you’re inviting near misses or worse. We almost had a nuclear exchange in the ’80s, and a Russian colonel saved the world by realizing the signals of an American strike were faulty. He refused to retaliate. That’s the kind of human judgment AI won’t be able to make.

Best case? A limited exchange—maybe each side fires ten missiles before someone slams the brakes. That’s 20 warheads, maybe 16 million dead in the first week, then hundreds of thousands more from fallout and cancer. That would be the worst war since WWII, but not extinction. Still, catastrophic.

Would that lead to eliminating nukes? Doubtful. No weapon in history has been retired without a replacement. Nuclear weapons will persist, now paired with AI. Treaties reduced U.S. and Russian warheads from 7,000+ each to around 1,700, but that’s still apocalyptic firepower. After losing 20 million, both sides cut to 250 each. Safer? Maybe. But I don’t know.

What I do know is we’re in a dangerous period. Not yet hazardous to the end of the world, but harmful. Climate change is slow-burning doom; nukes are instant doom. Both sit on our doorstep.

Jacobsen: Anything new in Alien: Earth world?

Rosner: They really have bad containment beakers. 

Jacobsen: How?

Rosner: They need to do reckless things; otherwise, you wouldn’t have a show, but the creature pops out. At this point, it is like a snake. It does not have arms, just a tail. It has its regular alien head. It is about five feet long from tip to tail. It can stand up on its tail, and it chitters at Wendy, who chitters back. She actually starts the chittering, and they begin communicating. The chest-burster finds Wendy’s chittering soothing, and she can reach out and pet it. That is the end of the episode.

We see things not shown in the movies. New elements include Wendy showing affection for the creature, the possibility of communication, and the chance that they can be reasoned with. In the films, Xenomorphs are typically portrayed as relentless killers and reproducers, using hosts to incubate offspring; the specifics of their diet remain ambiguous.

This friendship-like interaction between a young Xenomorph and Wendy is new to the canon in the TV continuity. Showrunner Noah Hawley has said he does not intend for Wendy to have a “pet” Xenomorph, but he is exploring the implications of limited communication.

Jacobsen: Is it a good or bad move for plot and canon? 

Rosner: It reads as a good move because it expands what can happen with them while keeping the horror tension—trust is fragile and likely to break.

Kirsh is Timothy Olyphant’s character, a synthetic mentor connected to the Prodigy corporation. Morrow is a cyborg adversary. The “Lost Boys” are six terminally ill children whose minds were transferred into adult synthetic bodies; Wendy is one of them. Boy Kavalier leads Prodigy. Joe Hermit is Wendy’s human brother. Dame Sylvia and Arthur Sylvia are married scientists.

Once chaos hits, we will probably see multiple species and threat types compounding—very much in the spirit of the original film’s life-cycle staging, where each phase (facehugger, chest-burster, adult) delivers a distinct kind of dread.

Alien looked convincing in 1979 and drew on H. R. Giger’s biomechanical, often erotically tinged designs—gooey, disturbing, and persuasive on screen. Star Wars (1977) reset visual expectations for sci-fi spectacle, while 2001: A Space Odyssey (1968) looked great but was neither horror nor a space western.

Earlier 1950s creature features often looked cheaper. The Thing from Another World (1951) featured a plant-based humanoid alien (famously derided as “carrot-like”), and The Blob (1958) centred on a rolling, amorphous mass—effective for the era, but less sophisticated than Alien’s design and staging.

Android “blood” in the Alien universe is a milky white fluid; synthetics can continue functioning after severe damage (as seen with Ash and Bishop in the films), which maps onto the likelihood that Prodigy’s hybrid bodies also use white coolant.

The question is how the humans will get hurt, and how the synthetics will be damaged. There are too many of them for the story not to push them into harm, and the damage will need some novelty. Out of the five, one has already had a nervous breakdown and is locked down. Given their strength, if she wants to get out, she can. The question is whether she stays unstable and whose side she will be on. By episode seven, you can predict there will be partially dismembered synthetics fighting, probably on the good side.

If the series stayed true to the films, nearly everyone would be killed. But this is television, and the creator has said that telling a TV story means telling it in 30, 50, even 70 hours. That means he expects multiple seasons. So not everybody will die.

The mayhem should still be perverse and satisfying, especially since capable people are running the show.

Jacobsen: Switching gears—Donald Trump has not been visible in public for almost a week. We have only seen him a few times, captured with telephoto lenses from far away. 

Rosner: The images are blurry, but he looks unwell. Reports say he went to his golf club in Virginia, but all we saw was him waving from a limo in a distant shot.

People are speculating about his health, noting he has never gone this long without speaking publicly or being seen. It looks convincing that he has had a health crisis. How serious it is, no one knows. He is scheduled to announce the Oval Office tomorrow afternoon.

Some claim he will resign, but that is unlikely. More realistically, it will be a policy announcement. He may send the National Guard into Chicago or issue an executive order on elections. 

Conducting elections is the responsibility of the states, aside from basic federal requirements, so any federal decree banning vote-by-mail, for example, would likely be challenged in court. The reasonable commentators expect a policy announcement, not a health update. Still, he has looked unhealthy—bruising on his right hand where IV needles could go, swollen ankles, and difficulty walking straight. We will see.

I have been thinking about this a lot. Trump has talked about running in 2028. If he is physically able, he will, because he loves the tens of millions in campaign contributions. He siphons money to himself, even though it is illegal. As long as people believe he is politically viable, ordinary supporters will donate, and wealthy donors will contribute because he can deliver for them politically. So he will announce he is running.

Everyone will say his run is unconstitutional. He will say, “Prove it.” People have worked out scenarios where he might try to stand against the Constitution or argue for a constitutional amendment. It is the 22nd Amendment that limits presidents to two terms, not the 25th.

He will continue to run, and cases will proceed through the courts. Different states will have different policies about whether he can be on the ballot. By June, July, or August, he will likely realize he cannot run and win.

One possibility—if no candidate gets a majority of electoral votes—the decision goes to the House, where each state gets one vote. Since there are more Republican states than Democratic ones, they could attempt to re-elect him. But that is extremely unlikely and unconstitutional.

More plausibly, by mid-2028, he will abandon the idea of another presidential run and instead try for a Senate seat in a red state. Florida gave him 56 percent in 2024, but states like Idaho gave him 66 or 67 percent. In a place like that, he would have a strong chance. As a Senator, he would keep pulling in campaign contributions and maintain influence. If Republicans take the Senate, he might even angle for majority leader. The money will continue to flow as long as he remains in power.

For that to happen, though, he has to keep functioning. If his brain declines, it barely matters to his followers—they like him no matter what he says. Everyone else is appalled, regardless. But some technologies could help him.

One is transcranial magnetic stimulation. It has shown benefits even in people with early Alzheimer’s. It is non-invasive: you wear a headset, a magnetic field passes through your brain, and it boosts neural activity. It sharpens brain function for approximately one to two hours afterward. If I were on Trump’s team, I would “juice” his brain with this before every public appearance.

In my near-future book, I imagine this happening—along with increasingly invasive interventions. My main character runs a neuro lab that is paid to keep Trump functional, eventually even “marionetting” him.

Elon Musk’s Neuralink has been tested on animals and some humans, allowing them to move objects with their thoughts. The concept could be reversed: implant devices in Trump so operators could control him, weekend-at-Bernie’s style. Assistive leg armatures already exist—they look like the exosuit Sigourney Weaver used in Aliens. Smaller, subtler versions could be developed. Or they could admit openly that he needs help walking.

Either way, the technology exists to keep him functional even as he declines. There are many possible interventions, because for his backers, the goal is simple: keep the money train running.

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.