Sep 25, 2024 · 1h 6m · bg2-pod
Ep16. Nuclear Update, AI Fast & Furious, State of VC | BG2 w/ Bill Gurley & Brad Gerstner · Bg2 Pod
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Co-hosts Brad Gerstner and Bill Gurley examine critical shifts across tech and finance, detailing nuclear power adoption for AI data centers, next-gen AI reasoning models, venture capital structural challenges, and macro stock market valuations.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Brad and Bill hold 99.2% of the talking time here. How this is scored →
Brad and Bill as informed peer 8.2 Guest teaching 1.7 Guest disagreement 1.5 Brad and Bill pushing back 2.3
speaking balance: gold is Brad and Bill, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 50:25 Gurley challenges institutional endowment unicorn marks
Gurley bluntly dismisses optimistic private valuations, arguing that 80 percent of zombie-corns are held on institutional LP books at unrealistically inflated prices.
Hardest push from Brad and Bill ▶ 46:39 Pushback against quasi-public categorizationGurley refuses Gerstner's premise that late-stage rounds equal public markets, arguing that absence of liquidity and public market discipline fundamentally distorts execution.
Biggest teaching moment ▶ 29:41 All revenue is not created equal in AI computeGurley educates on gross margin realities, explaining that heavy GPU token consumption makes AI economics fundamentally different from high-margin ad businesses.
Brad and Bill hold their own ▶ 57:33 Invoking the observer effect in venture capitalGurley halts the wrap-up to deploy the scientific concept of the observer effect, arguing late-stage mega-capital fundamentally perturbs and damages startup execution.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Brad and Bill as informed peer | Guest teaching | Guest disagreement | Brad and Bill pushing back | Why |
|---|---|---|---|---|---|---|
| Nuclear Renaissance and Hyperscaler Energy Demand Update | 8 | 1 | 0 | 1 | Bill Gurley and Brad Gerstner discuss nuclear energy for AI hyperscalers, referencing Moore's law analogies, carbon offset market math, and Crossing the Chasm frameworks in a highly collaborative peer exchange. | |
| AI Fast & Furious: Infrastructure Demand and Capital Expenditure | 8 | 1 | 1 | 1 | Both hosts evaluate AI data center capex commitments, with Gurley highlighting new multi-billion dollar financing vehicles from BlackRock, Microsoft, and Middle Eastern sovereign funds. | |
| Next-Gen AI Models, Inference Scaling, and User Growth | 8 | 2 | 1 | 2 | The hosts break down inference scaling laws, OpenAI's o1 model reasoning trade-offs, and economic compute constraints, citing logarithmic performance graphs. | |
| OpenAI's Strategic Moat, Flywheel Network Effects, and Valuation | 8 | 2 | 2 | 4 | Gurley pushes back on Gerstner comparing OpenAI directly to Google and Meta valuations, pointing out that AI gross margin structures may resemble AWS or airline fuel variable costs rather than pure software. | |
| State of VC: Startup Overfeeding, Secondary Liquidity, and Market Dynamics | 9 | 3 | 4 | 5 | A spirited debate on the state of venture capital where Gurley argues mega-funds overfeed startups and suppress IPOs via secondaries, challenging Gerstner's quasi-public framing. | |
| Macro Tech Check: Fed Rate Cut Impact and Market Valuations | 8 | 1 | 1 | 1 | Gerstner leads the macro assessment of the Fed 50bps rate cut, dissecting Mag 7 forward PE multiples versus growth PEG ratios while Gurley probes consumer health. |
Statements from this episode (16)
Gurley: Hyperscaler buyers are 10x better for nuclear startups than utilities
“If the hyperscalers become part of the customer set for the nuclear startups, that may be like 10 X better than selling just to utilities alone. Like you may have brought innovators to the table. On the purchasing side that may be more open-minded, that may be…”
Gerstner confirms Altimeter is in talks to join OpenAI's funding round
“I can confirm that Altimeter is talking with the company, so of course there's some things I can share and things I can't share”
Gurley: AI compute spending will shift from training toward inference over time
“Maybe the dollars of compute are going to move more towards inference than training as we move forward.”
Gerstner: The AI world is currently constrained by inference compute
“I think the world is inference constrained. I think part of the reason for that is that you have new models emerging like this that are going to have machines talking to machines, lots of inference going on in the background.”
Gerstner: ChatGPT is pulling away; Gemini, Meta AI, and Claude are lagging
“ChatGPT beginning to run away with it. Gemini, Meta AI, Claude really are not even you know, keeping up.”
Gurley: Integrated personal data creates insurmountable switching costs for AI assistants
“If I have an AI partner where I can simply say, who did I send that email to? Like, that's really, really powerful. And I think the switching costs are insurmountable if someone gets to that place first.”
Gerstner: AI inference costs have fallen over 90% in 18 months
“We already know the cost of inference has come down by over 90% over the course of the last 18 months, and our friend Sonny over the weekend when we were in our group chat, you know, said he expects it to come down by another 90% over the course of the next se…”
Gurley: Funding founder secondaries eliminates the primary incentive for tech IPOs
“And if you're going to be founder friendly and write big checks, guess what? You're going to be supportive of founder secondary. And you're going to be in for, you're going to be supportive of broad based employee secondary. When you do those things, you are t…”
Gerstner: 80% of current unicorn startups will never regain peak valuations
“Lots of people talk about the zombie corns, right? We have a thousand companies that are, that were unicorns. A lot of those will never get back there. I've said, 80% of those companies will never get back there.”
Gerstner: Altimeter expects four portfolio IPOs, including Cerebras, CoreWeave, and Databricks
“I think we may have four IPOs in the pipeline in the next four to six months. The rumors out there around companies like Cerebris and CoreWeave and Databricks, those are all in our portfolios.”
Gurley: Large endowments are holding zombiecorn startups at unrealistically high valuations
“And by the way, you talk, if 80% of the zombie corns are never going to get out or are going to get out at reduced prices, I'm telling you flat out, those are being held on the large endowments, you know, as LPs at unrealistically high prices across the board.”
Gerstner: Altimeter pushed to sell Lacework despite having hundreds of millions remaining
“We had a company that wasn't performing at the levels that you know, it previously had been. It had been priced at many billions of dollars in, you know, at the peak of Zurb in 20, 21, a company called Lacework. And we ultimately, you know, we pushed to sell t…”
Gerstner: Unnecessary $500M mega-rounds trigger secondary demands and erode employee retention
“You can't go raise another five hundred million and not have pressure from all your employees for employee secondary For spending more money on more projects, et cetera, and the NPV on those other activities will be lower, and the incentive your employees have…”
Gurley: Requiring $500M revenue for IPOs will create excessive zombiecorns
“If that becomes true, if that kind of Path dependency is, is cast upon every venture capital company that comes along. You're going to end up with an excessive amount of zombie corns because previous to this evolution in the late stage markets, plenty of compa…”
Gerstner: The entire Nvidia and AI infrastructure supply chain remains undersupplied
“We think the entire NVIDIA and AI infrastructure supply chain is going to continue to be under supplied.”
Gerstner: Nvidia's 2025 stock hinges on shipping 6M versus 4.5M GPUs
“You know, we're kind of at, You know, six million GPUs for next year. The bearish people are at like four and a half million. Like the numbers will ultimately tell if they do six million next year, the stock's going higher. If they do four and a half million, …”