OP #309: The Loop: What Nvidia and OpenAI Teach Us About Circular Investing

Some good news, life skills vs. academic skills, and circular deals. Wow, that's a lot to cover. Enjoy the OP.

Author

Darren Herman
September 30, 2025

Sitting at RDU on a 5.5-hour delay (oh well). This weekend our oldest committed to High Point University—Class of 2030. Eighteen months ago I hadn’t heard of HPU; since then I’ve come to appreciate how clearly they’ve positioned the place and where they’re headed.

I’ve learned a lot about education marketing through Tremont School—a small but mighty school for neurodiverse students near home. On college tours with my son, most pitches blurred together. Same brochures. Same talking points. Same promises.

High Point stood out.

Yes, they focus on academics. They also make life skills a core track, and they point to outcomes (they cite a 99% job placement rate within six months of graduation). It’s not just what you learn—it’s the habits you leave with.

I still wrestle with the ROI of a traditional degree. At around $85k a year, you’re staring at a ~$340k decision before aid—often to reach table stakes in business. Advanced degrees used to be the differentiator; with AI in the mix, the calculus is shifting.

Let me break this down again. $340k in cash before financial aid. Lets assume you earn 5% on this money (conservative) and compound it over a 10 year period, it’s actually $553k. So, you’re not only sacrificing the principal of $340k but you’re giving up $213k in potential [conservative] interest or investment related returns. Could that money be better invested elsewhere?

And you’re now potentially $553k in the hole and competing with not only more college graduates than ever, but now… AI which will likely be academically smarter than you (and me) will ever be.

Trade schools may have a tailwind—they’re closer to the work and less exposed to near-term automation. Worth more attention than they get.

We chose High Point because it aligns with our view: prepare students for life beyond the classroom and help them become their best selves. Their tagline—“the premier life skills university”—reads less like a slogan and more like an operating principle. If AI is augmenting academic skills, human skills matter even more. This is Sherri and I’s thinking… hopefully it’s right.

Also: their marketing is strong. The team behind it is one to watch.

Music break: Spotify nailed a mostly-rock mix—Oasis, Weezer, The Killers, Jet, The Strokes, Blur, and more—that’s been fueling my gym sessions. Listen here.

Here are five articles to get us started with:

Happy Tuesday. I hope you are having a wonderful week. If someone forwarded you the OP and you’d like to subscribe for free, you can do so here.

Be well, do good.
Darren

The Loop: What Nvidia and OpenAI Teach Us About Circular Investing

Ever since I heard Ed Edelson (and Gil Luria) break down the OpenAI–Nvidia deal, I haven’t been able to stop thinking about circular investments. Ed is Scott Galloway’s partner on the podcast, and frankly, I tend to like him a lot. He’s sharp, measured, and has this way of connecting dots without overhyping them.

The idea he surfaced—and that I’ve been chewing on since—is what’s called a “circular investment model.” Not the sustainability kind you see in ESG decks, but the way money itself loops between two players, reinforcing the partnership in real time.

Traditional investing is linear: you put money in, wait, and eventually (hopefully) see a return. Circular setups are different. Capital cycles back almost immediately. One side invests, the other spends that capital on the investor’s products or services, and the loop reinforces itself.

That’s exactly what’s happening between Nvidia and OpenAI.

OpenAI is chasing AGI, which demands an absurd amount of compute—millions of GPUs, entire data centers built on silicon muscle. Nvidia makes those chips. And now, Nvidia isn’t just the supplier, they’re also the investor.

Here’s the loop:

  • Step 1: The Investment. Nvidia pledges up to $100 billion into OpenAI, structured to grow as OpenAI grows.

  • Step 2: The Purchase. OpenAI turns around and spends that money buying Nvidia’s chips. The capital leaves, and then it comes right back.

For Nvidia, this locks in a long-term customer and ties their fortunes to one of the most ambitious AI projects ever. For OpenAI, it guarantees access to the hardware it needs to even have a shot at building AGI.

It’s a fascinating model because it collapses the lag between investment and return. Both sides are betting not just on the future, but on the momentum of their partnership today.

But here’s why I’ve been dwelling on it: I also see the risks.

I had a sideline seat to the original dot-com boom and now, an active role in this one. The crash was brutal—but it didn’t kill the internet. Quite the opposite. The web grew back stronger, healthier. I wonder what version of that story plays out for AI.

Three risks stand out to me:

1. Systemic Fragility.
When everything leans too hard on one player, the whole system wobbles. If OpenAI stumbles—whether from competition or underperformance—Nvidia loses twice: once on its investment and again on chip sales. That’s concentration risk at its ugliest, and it doesn’t stop at two companies. The ripple can shake the entire sector.

2. Bubble Logic.
Circular flows prop up asset values beyond reality. Nvidia chips look like they have endless demand, valuations climb, investors pile in. Until they don’t. Bubbles never ease down; they pop.

3. The Illusion of Growth.
At its worst, circular investing resembles recycled debt. Dollars spin, activity looks busy, value looks real—but it’s not creating anything new. We’ve seen this movie before in financial crises. Shaky foundations don’t hold.

As someone deeply fascinated by this ecosystem—and yes, bullish on what the next few decades of AI could bring—I want to see long-term, sustainable investing. I’d like to think we learned something from the late ’90s crash. Let’s hope we did.

Below are a few articles I came across this past week that I found interesting. While I may not agree with everything in each one, I think they're worth a read. If you stumble upon an article you think I or the Operating Partner community would enjoy, feel free to share it with me. Of course, I reserve the right to decide what gets featured in the OP.

Thank you again for subscribing to the OP. I had a lot of fun writing this one and hope you enjoyed it.

What did you think of today's OP letter?

Login or Subscribe to participate in polls.