Episode summary

$6 Billion Investor on Data Centers, AI & the Next Trillion-Dollar Companies

How I Invest With David Weisburd · · 48 min · Hosts: David Weisburd · Guests: Rob Stephens

$6 Billion Investor on Data Centers, AI & the Next Trillion-Dollar Companies – How I Invest With David Weisburd

Summary

Rob Stephens argues that power, not capital, is becoming the key constraint in AI and technology, which is pushing investors to blur the old boundaries between venture capital, natural resources, and infrastructure. He says data centers, electrification, and even space-based computing are increasingly connected, because the biggest AI and industrial-tech businesses now depend on massive electricity demand and new ways to supply it. That shift, in his view, makes traditional asset-allocation buckets feel too rigid for how opportunities actually show up.

He also says being both a former GP and current LP gives him a sharper lens on manager behavior, especially around performance presentation, alignment, and co-investments. Stephens prefers managers whose incentives are clear and whose track records can be verified through references, not just marketing materials. He views SPVs, co-investments, and continuation vehicles as useful tools, but ones that can distort portfolios if they are used to chase fees, obscure weak assets, or overextend concentration.

On venture capital, Stephens thinks the market is increasingly split between elite firms with access to the biggest private winners and everyone else. He expects the best late-stage companies to keep accruing value privately longer, but he also thinks that creates a cycle where private markets gain more importance, more capital, and eventually even more staying power.

Key takeaways

  • Rob Stephens says power is the main bottleneck for AI, making data centers, energy infrastructure, and natural resources central to future private-market investing.
  • Stephens believes the private market is increasingly dominated by a handful of elite firms with access to Anthropic, OpenAI, and SpaceX-type outcomes.
  • He says LPs should scrutinize GP incentives closely, especially when co-investments, SPVs, or continuation vehicles can change economics and obscure track records.
  • Stephens argues that references matter more than portable track records when evaluating spinout managers because venture outcomes are still a team sport.
  • He thinks conventional asset-class buckets are too rigid for thematic strategies that sit between venture, credit, equity, and natural resources.

Notable quotes

“Power is the main bottleneck for everything that the US is trying to do in the AI space and in technology broadly speaking”

— A speaker on How I Invest With David Weisburd

“There are ways to to massage the numbers and also to tell a story that befits the GP more than perhaps the LP knows if they're not kind of aware”

— Rob Stephens

“We are seeing a concentration of winners in the latest stage private growth companies and private tech in particular that are creating a halves and have nots of LPs and investors”

— Rob Stephens at 4:11

“References that is the holy grail in terms of trying to figure it out”

— Rob Stephens at 24:23

“The thesis broadly is that demand for AI and new technology and electrification here in America is insatable”

— Rob Stephens at 47:52

How the episode opens

Power is the main bottleneck for everything that the US is trying to do in the AI space and in technology broadly speaking. We have data centers in Richmond, Virginia out by the airport that I didn't even know existed. I drive by there all the time to go home to my childhood town and these massive Meta and Amazon data centers are just hidden back there behind the trees.

You get back there and you say like, "How big is this building? And they don't answer you in square footage. They answer you in megawatt.

Rob, you're director of investments at Spider Management, which has over 6 billion. And before you were an LP, you were also a GP for 10 years. How does that change how you approach being an LP?

I think honestly it provides a little bit of a different lens than a lot of my peers have because they don't have the experience having sat on the other side and recognizing some of the nuance to the way that GPS speak to LPS some of the things that might get hidden behind the marketing materials behind the return figures that are presented. There are ways to to massage the numbers and also to tell a story that befits the GP more than perhaps the LP knows if they're not kind of aware. and being on the LP side now for half a decade, what are some of these patterns?

What are some things that GPS are doing that may not be apparent to LPS?

More subtle things. It's not always in writing. It's not always something that you can find in the materials. Perhaps the easiest one is just recognizing whether or not the GPS are putting their performance. If they have it, if the fund is or firm has been around long enough, then the performance should be front and center or at least available.

From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →

Topics

  • ai reshaping capital allocation
  • data centers and energy infrastructure
  • venture co-investments and spvs
  • private markets versus public markets
  • concentration vs diversification

Companies and tickers mentioned

  • Spider Management
  • DAC Management
  • Anthropic
  • OpenAI
  • SpaceX

Watch the episode

Watch “$6 Billion Investor on Data Centers, AI & the Next Trillion-Dollar Companies” on YouTube

Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.

More from How I Invest With David Weisburd

All How I Invest With David Weisburd episodes →

How I Invest With David Weisburd, this episode and its artwork belong to David Weisburd. Octagon is not affiliated with the show; this page is an independent summary.