Rick Rosner On, 31: AI, Education, Einstein, Gamow, and Mathematical Intuition

Author(s): Scott Douglas Jacobsen

Publication (Outlet/Website): Personal SubStack

Publication Date (yyyy/mm/dd): 2026/08/23

In this interview, Scott Douglas Jacobsen and Rick Rosner discuss whether mathematical intuition or formal math skills matter more in modern life. Rosner examines South Korea’s education culture, AI-assisted computation, elite university STEM trends, and the value of recognizing bad outputs. He also reflects on Einstein and Marcel Grossmann, George Gamow, Hardy-Weinberg equilibrium, and why imagination, collaboration, practical knowledge, and mathematical intuition may become increasingly important as AI handles more technical work in everyday society.

Math Sense, Math Skills, and South Korean Education

Scott Douglas Jacobsen: What’s more important to get through life in most societies? A good math sense or good math skills?

Rick Rosner: It depends on the society. For instance, I’ve been watching a show called Get Schooled from South Korea.

We know that South Korea has an intensely competitive education system, right? There’s enormous pressure surrounding academic performance and the College Scholastic Ability Test, or CSAT, which plays an important role in university admissions. There’s also a huge private tutoring industry, including hagwons.

It’s a whole thing. Get Schooled is a South Korean webtoon about abuses within the education system involving parents, teachers, students, and others. In the story, the government creates an Education Rights Protection Bureau, whose agents intervene in troubled schools. It’s fictional, not a depiction of an actual South Korean government bureau.

It’s a different type of culprit in different story arcs, but the agent finds the bad people and fucks them up.

The “Fellow Kids” Meme and the Undercover-Student Story

Jacobsen: Hello, kids. For those who don’t know, there’s a funny Steve Buscemi scene with a hat on backwards and a skateboard. He’s an undercover student, way, way too old. There’s this meme, “How do you do, fellow kids?” That’s Steve Buscemi. So he looks even older and weirder.

Rosner: And he’s wearing a band shirt, but the shirt says “Music Band.” What’s weird is, two hours ago, it popped up on my Twitter feed. Carole was in the room.

I go, “Hey, look, that’s me in high school,” because I was a way-too-old undercover student.

Jacobsen: You were. I don’t know when the Steve Buscemi thing was, but I’m pretty sure it was after the Rick Rosner literal undercover-cop student. Then you’re getting over that.

I had quite a few. Because you had the physics equation on that plane, the crop plane. Then you met Carole at your other thing, going to a Jewish singles dance. And then you two met.

Carole and Rosner’s Eccentricities

Rosner: And then I decided Carole was worth being with, because she went along with me on this stupid idea of pretending to be in high school.

Jacobsen: It’s your greatest strength and your greatest weakness, your honesty. You go too far sometimes. But when it came to something as serious as being an undercover student for your own reasons, you telling her this and her being tolerant was a good thing, because you found someone who would be willing to tolerate your eccentricities.

Rosner: Yes, right. We’ve talked about it recently, this past week. She attributed it mostly to her being dumb and naive. She hadn’t had that many…

Jacobsen: She’s not a dumb person. She is not a naive person.

Rosner: Yes. But in this instance, she hadn’t had that many boyfriends. She went along with this thing.

Screenplay Ideas and the Proactive Heroine

It seemed harmless, right? I want to write a screenplay with her where a guy like me, but even worse, gets her involved because of her naivete. That’s the setup to several movies where a bad boyfriend drags an innocent woman into a bad drug situation. But in the screenplay, I envision the girl may be naive, but she’s seen a ton of movies and TV shows.

So when it comes to the point where they’re in the bad couple’s apartment, and some dire shit is about to go down, instead of waiting for it to become dire, Carole just shoots the guy. The dumb boyfriend is, “What? Why did you shoot the guy?”

And she goes, “You knew where that was going.” She shoots the guy. She holds the bad guy’s female partner at gunpoint, maybe ties her up, maybe pistol-whips her, and saves the dumb boyfriend without waiting for the usual mechanics we’ve all seen, the increasing tension in the scene and the narrow escape. Instead, this naive but not stupid young woman is, “Fuck this. I see what this is. I’m just going to shoot him right now.”

We also have another screenplay that works the same way, unofficially titled, working-titled, because we haven’t written it and I want to, Carole the Office Assistant, where you have a character who’s proactive with her violence because she’s not going to wait for the shit to play out. She’s seen it 50 times in a movie.

She’s going to shoot somebody before they can shoot her, which would be fun. It would move faster. But anyway, back to South Korea.

Mathematics, Elite Education, and AI

South Korea is math-heavy. China is math-heavy. Surprisingly, the U.S. is freaking math-heavy too, particularly at highly selective universities. At Harvard, for example, just over half of the undergraduate degrees awarded in the 2024-25 academic year were in engineering, applied sciences, or science. You can’t generalize that exact percentage to every Ivy League university, though. Yale’s recent figures, for example, put STEM at roughly a third to somewhat under 40 percent of bachelor’s degrees. So the idea that these universities are mostly people studying some useless political bullshit is wrong. Huge numbers of students are doing science, engineering, computer science, mathematics, and other quantitative work.

Five years ago, say, I’d have said actual mathematical knowledge might be more important. But now AI is coming in and can do a lot of the computational and coding work. I would shift some emphasis back toward having an intuitive understanding of math, because you can use AI to help with calculations and formal work. You still need enough knowledge to recognize when the AI is bullshitting you.

I didn’t have to figure out all the math of the fucking golf balls lipping out. I could say, “Hey, AI, describe the physics of a golf ball lipping out.” You’re almost falling asleep.

But anyway, it’s better to have both, right, up to a point. You should have basic math knowledge and as much advanced math knowledge as is useful to you. Advanced mathematics doesn’t inherently stifle imagination. For some people, it gives them more ways to express it.

Einstein, Grossmann, and Mathematical Collaboration

A major example is Einstein developing general relativity. He was a brilliant physicist, but he needed mathematical machinery he hadn’t previously mastered. He first arrived at the key physical insight behind general relativity in 1907 and completed the theory in 1915, so the development took about eight years rather than 10.

And what he had was a mathematically sophisticated friend, Marcel Grossmann. This wasn’t in Vienna, and it wasn’t during the final stages while Einstein was working at the Swiss Patent Office in Bern. By 1912, Einstein was back at ETH Zurich as a professor, and Grossmann was a mathematician there. Grossmann introduced him to differential geometry and tensor calculus, including the work of Riemann, Ricci, and Levi-Civita, which became essential to general relativity.

Grossmann didn’t suddenly hand Einstein a four-by-four matrix with 10 numbers and expect everything to work. The metric tensor in four-dimensional spacetime can indeed be represented as a symmetric four-by-four array with 10 independent components. Still, the mathematics involves tensor calculus, differential geometry, curvature tensors, and differential equations. Einstein and Grossmann actually explored some of the right mathematical machinery in 1912 and 1913, moved away from part of it, and Einstein returned to the crucial formulation in 1915.

So it’s important to have mathy friends.

George Gamow: Intuition, Collaboration, and Excess

Frickin’ Big Bang guy, George Gamow. He wasn’t terrible at math. You couldn’t make his contributions to quantum tunnelling, nuclear physics, cosmology, and Big Bang nucleosynthesis if you were mathematically incompetent. But he was famously intuitive and imaginative, and he sometimes relied on collaborators who were stronger in particular areas of mathematical calculation.

He joined the University of Colorado Boulder in 1956. There’s a physics tower there named Gamow Tower. He collaborated with people who could work out the mathematical details of ideas that interested him.

He was a party guy. He was a drinky guy. He had a good time. He was a heavy drinker, and accounts of his later life say alcohol damaged his health and professional effectiveness. He died in 1968 at 64 after serious health problems, including liver failure. So it’s reasonable to say his drinking may have contributed to his early death, but you can’t reduce the cause of death simply to drinking.

While he lived, though, he was extraordinarily imaginative and fun. Plenty of stories and photographs reflect that side of him.

He started in the Soviet Union as a young man and then got the fuck out. Meanwhile, he had fun even under a repressive regime. Yes.

Hardy-Weinberg Equilibrium

What else? There’s also the Hardy-Weinberg equation in population genetics. Reginald Punnett, the geneticist, went to G. H. Hardy, one of the most famous mathematicians of the early 20th century, with a question about Mendelian inheritance in populations. Punnett wanted to understand why, under random mating, a dominant allele wouldn’t simply drive a recessive allele out of the population.

Hardy basically said, “Well, that’s simple.” If the frequencies of two alleles are (p) and (q), with (p+q=1), then under the standard Hardy-Weinberg assumptions the genotype frequencies after random mating are (p^2), (2pq), and (q^2). If the two alleles each happen to have a frequency of one-half, then the frequencies are one-quarter, one-half, and one-quarter.

Punnett later associated Hardy’s name with the result. Independently, the German physician Wilhelm Weinberg had derived essentially the same principle earlier in 1908, which is why we now call it Hardy-Weinberg equilibrium. It isn’t an equation of heritability. It’s a population-genetics model describing the relationship between allele and genotype frequencies under specified assumptions.

Hardy described the point as very simple and said he expected biologists to find it familiar. The later description of Hardy-Weinberg equilibrium as a result of “embarrassing simplicity” comes from historians of genetics, so saying Hardy himself was pissed off or embarrassed by having his name attached to it goes beyond what we can establish.

Mathematical Intuition and the Future of AI

Anyway, better to have mathematical intuition tempered with some actual mathematical knowledge. The best way is to live 10 or 20 years from now, and maybe you’ll have some brain-computer interface that links you to whatever mathematics you need through AI. Whether anything remotely that capable exists on that timetable is anybody’s guess. The early versions will probably be shitty.

Jacobsen: Thank you very much for the opportunity and your time, Rick.

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.