Jan 11, 2025 · 1h 23m · bg2-pod
Market Predictions, Rates & Inflation, DOGE, CES, AI Compute | BG2 w/ Bill Gurley & Brad Gerstner · Bg2 Pod
Brad Gerstner · 51m spoken Bill Gurley · 24m spoken Jensen Huang · 12s spoken Elon Musk · 11s spoken
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
In this episode of BG², venture capitalists Brad Gerstner and Bill Gurley analyze the 2025 macroeconomic landscape, detailing the multi-billion-dollar AI infrastructure CapEx surge, evolving model scaling laws, federal budget reform with DOGE, and state-versus-federal tech regulation.
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 99.4% of the talking time here. How this is scored →
Brad and Bill as informed peer 7.8 Guest teaching 0.0 Guest disagreement 0.1 Brad and Bill pushing back 1.0
speaking balance: gold is Brad and Bill, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 26:55 Brad dismisses PR-driven CapEx framing
Brad rejects the notion that hyperscalers announce CapEx merely for optics, directly clarifying that real physical compute constraints dictate their investments.
Hardest push from Brad and Bill ▶ 58:08 Bill blasts state-level AI regulationsBill aggressively criticizes state-by-state AI legislative initiatives as economically damaging red tape and regulatory capture.
Biggest teaching moment ▶ 38:20 Brad explains HBM memory valuation mispricingBrad explains why the public market misprices SK Hynix at a 6x multiple as a cyclical commodity instead of secular high-bandwidth memory.
Brad and Bill hold their own ▶ 10:18 Brad dissects Mag 5 vs S&P 495 earnings disparityBrad provides granular financial data demonstrating that the Mag 5 grew earnings 44% while the remaining 495 companies grew only 2%.
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 |
|---|---|---|---|---|---|---|
| LA Wildfires, Governance, and Public Accountability | 6 | 0 | 0 | 0 | Brad and Bill discuss the LA fires and the disconnect between policy intent and actual outcomes. Both align closely on accountability, sharing thoughtful perspectives without confrontation. | |
| 2025 Annual Investment Framework and Year-End Planning | 7 | 0 | 0 | 0 | Bill prompts Brad to explain Altimeter's year-end process. Brad articulates how his fund balances macro forecasting, journal reflections, and public versus private market horizons. | |
| Macroeconomic Backdrop: Valuations, Inflation, and Yields | 7 | 0 | 0 | 0 | Brad lays out the macro backdrop for 2025, contrasting tailwinds like deregulation and AI against potential risks such as elevated valuations and rising 10-year Treasury yields. | |
| Big Tech Valuations, Earnings Growth, and Non-Tech Expectations | 8 | 0 | 0 | 2 | Brad cites specific multiple compressions and earnings data for Mag 5 versus the remaining S&P 495. Bill interjects to note the historic anomaly of their growth at current scale. | |
| Big Tech CapEx Surge and AI Infrastructure Investment | 8 | 0 | 0 | 1 | The hosts break down hyperscaler CapEx stepping up to 260 billion dollars. They weave in clips of Jensen Huang discussing the exponential increase in compute needed for test-time reasoning. | |
| Evaluating ROI on AI Infrastructure and Labor Automation | 8 | 0 | 0 | 3 | Bill questions whether spending 100 percent of free cash flow on CapEx is prudent. Brad explains that early-stage phase shifts defy linear Wall Street modeling, citing labor automation upside. | |
| Hyperscaler Competition, Monetization, and Compute Bottlenecks | 8 | 0 | 1 | 2 | Bill queries whether CapEx guidance is driven by perception reflexivity and prisoner's dilemmas. Brad rejects this framing as overly conspiratorial, citing severe real-world compute bottlenecks. | |
| Three Scaling Laws: Test-Time Reasoning and Agentic Compute | 8 | 0 | 0 | 0 | Brad analyzes how test-time compute, deep reasoning, and personalized agents create a massive multiplier on token generation and infrastructure demand relative to single-shot queries. | |
| CapEx Winners: Semiconductor Complex, Memory, and Energy Constraints | 9 | 0 | 0 | 2 | Bill and Brad evaluate derivative winners from CapEx, examining Nvidia data center revenue, SK Hynix HBM memory multiples, and nuclear energy constraints like Diablo Canyon. | |
| Interest Rates, Treasury Yields, and DOGE Deficit Reduction | 7 | 0 | 0 | 0 | Brad examines why 10-year yields rose despite falling headline inflation. He argues the market doubts Congress will enact offsetting spending cuts against anticipated tax stimulus. | |
| Federal Budget Breakdown and Fiscal Reform Possibilities | 8 | 0 | 0 | 1 | Brad reviews historical federal spending charts, showing a 1.7 trillion gap above baseline growth. Bill adds that structural healthcare and demographic costs require targeted reforms. | |
| AI Regulation: Federal Preemption vs. State-Level TRAIGA | 9 | 0 | 0 | 1 | Bill strongly condemns state-level AI regulation bills like Texas TRAIGA, detailing how compliance audits cause regulatory capture. Brad advocates strongly for federal preemption. | |
| The Future of Search: Google's AI Assets vs. Apple | 8 | 0 | 0 | 1 | Bill outlines Google's vertical advantages in consumer and enterprise AI applications. Brad counters by highlighting the innovator's dilemma of defending high-margin search revenue. | |
| xAI Colossus Cluster and Elon Musk's Three AI Vectors | 8 | 0 | 0 | 1 | Bill queries whether Colossus provides xAI an edge. Brad maps Elon's three AI vectors across robotics, full self-driving, and language models as an integrated keiretsu structure. | |
| Big Tech Co-opetition and Reasoning Model Frontiers | 8 | 0 | 0 | 1 | The hosts discuss hyperscaler coopetition and reasoning model scaling. Bill highlights the counterintuitive compute intensity of chain-of-thought models, while Brad forecasts agentic breakthroughs. |
Statements from this episode (34)
Gerstner: Inflation and Higher Rates Could Dampen Markets Early in 2025
“And I really think that it's this inflation and higher rates Combined with higher valuations that could put a damper on at least the market in the first part of this year.”
Gerstner: Big Tech Valuations Are Not in a Bubble
“I look at that and I say, does, is that a bubble? Right? Does this feel like a bubble? I say, no. It's not as good a deal as you got at the start of 23. And it's not as good a deal as you got to start at 24, but if we achieve the growth expectations we expect …”
Gerstner: Big Five Grew 2024 Earnings by 44% vs 2% for S&P 495
“And in 2024, the big five, so in this case, Microsoft, Meta, Nvidia, Amazon, and Google grew earnings at 44%... If you look at the other 495 companies in the S&P, their earnings only grew two percent.”
Gerstner: 2025 Market Returns Depend on Non-Tech Earnings and Sub-5% Rates
“I don't think we're going to see a year that looks anything like 20, 23 or 20, 24 anyway. But if you just said, can we get to 10 or 15% on the market? A couple of things have to happen. You have to see this acceleration in earnings, number one, and number two,…”
Gerstner: Microsoft is on an $80 billion annual CapEx run rate
“In their most recent quarter, they did about twenty billion in capex bill, and you just, you know, he said earlier in the year, which caused, or a couple of days ago, which caused a bit of a kerfuffle that they expected to spend eighty billion in capex this ye…”
Gerstner: Big tech AI CapEx investments will materialize in 2025
“And I think what we can put to bed based upon the conversations we've had, what people are hearing out there, the conversations at CES, there are going to, these investments are going to show up in 20, 25, right?”
Jensen Huang: AI inference demand will scale up by one billion times
“It's gonna go up a billion times.”
Gerstner: Apple is not investing in frontier AI models
“Because they're not, you know, Apple is not investing in frontier models.”
Gerstner: Microsoft generates $10 billion in AI inference revenue
“You know, as Satya told us on the pod, they have ten billion in inference revenue, and I think he expects that to grow. Significantly.”
Musk: AI will perform any cognitive task within three to four years
“I mean, AI really within the next few years will be able to do any cognitive task. Like, it obviously begs the question, what are we all gonna do? You know, I'd say max three or four years. Maximum.”
Gerstner: 2023 Wall Street consensus forecast for Nvidia missed by 80%
“In 20, 23, you had 26 analysts on Wall Street covering NVIDIA. And the consensus forecast of all 26 of them missed by 80%.”
Gerstner: Meta charges hyperscalers licensing fees for Llama models
“Where he said they already charge license fees to the large hyperscalers to use Lama and to provide it to them in certain ways, etc. I think it's de minimis revenue in the scheme of Meta, but, you know, he's smart enough in that. Bill Gurley: I don't know. Bra…”
Gerstner: AI will remain compute-constrained through the end of 2025
“And so I think the investments we're seeing in 20, 25 Frankly, I don't even think it's going to get us to the point where we have compute surplus. I think we're going to end this year compute constrained, and I think a lot of the frontier labs feel the same.”
Gerstner: Mistral and other model companies pivoted away from frontier spend
“Mistral, a lot of the other model companies have fully pivoted, like they've already raised their hand, white flag, we don't have the revenues that can support the spend to keep up, right?”
Gerstner: Nvidia inference revenue is upwards of 50% of total revenue
“Jensen made, and maybe we'll insert the clip here on the pod, where he talks about his inference revenue is already 50% of his revenue, so already upwards of 50% of his revenue”
Gerstner: Reasoning models dwarf single-shot ChatGPT in compute consumption
“You have these O series models, and now you have deep reasoning out of Google that launch this whole different vector of scaling intelligence and reasoning. But the thing about those, Bill, as you well know, is they are compute hogs. Right? The number of token…”
Gurley: Hyperscaler CapEx spending is sticky, slow, and irreversible
“The processes for spending that amount of money are sticky and slow. They're not fast. Like, you can't back out. Like, you pre-commit, and you go build.”
Gerstner: Nvidia will not lose AI market share from 2024 to 2025
“So I actually think if you look at the consensus numbers, they probably would imply that NVIDIA's market share is going down 24 to 25. And I don't think that to be true.”
Gerstner: High-bandwidth memory will face shortages for foreseeable future
“And we think that we have a memory shortage, which is a key part of the NVIDIA supercomputer ecosystem for as far as we can see.”
Gerstner: Newsom has not signed Diablo Canyon extension past 2029
“Gavin Newsom still has not signed the extension for Diablo Canyon beyond 2029. It's insane. We have to make that happen this year. It's 10% of California's green, clean power.”
Gerstner: AI power stocks are derivative bets; buy Nvidia instead
“And so I would always ask my analysts when they would bring me a power company, I would say, why is that better than NVIDIA? Right? It's a derivative NVIDIA. It's the exact same bet. So why not just buy more NVIDIA?”
Gerstner: DOGE and Trump will cut hundreds of billions from federal budget
“My forecast is that Doge is actually, and this administration are actually going to cut a lot out of the budget. And you know, I think that you have the leadership in the House and the Senate, and so if they cut hundreds of billions of dollars out of this budg…”
Gerstner: FY2024 federal spending exceeded baseline 2.5% growth by $1.7T
“And if you see there, the baseline, 2019 budget adjusted two and a half percent CAGR. We should have spent five trillion, but instead we spent 6.7 trillion. That's a 1.7 trillion dollar differential.”
Gerstner: Tax cuts and AI productivity will accelerate US growth
“Cutting taxes, three to four hundred billion dollars a year is going to accelerate it. We are going to get efficiency and productivity gains from AI. Like these things are happening.”
Gerstner: Trump will likely rescind most of Biden's AI executive order
“I suspect that Trump is going to rescind all, or at least part, the major part of the Biden EO executive order on AI.”
Gurley reports up to 25 state-level AI regulatory initiatives underway
“I've heard there's as many as 25 now state-by-state initiatives, and I guess this is how policy works in this country, but there's a bunch of reasons to be really worried about this.”
Gerstner: Significantly more AI companies will open-source models in 2025
“I think you're going to see in 20, 25, a lot more companies open source more of their models.”
Gurley: Google possesses a consumer AI asset stack Apple and Microsoft lack
“And, you know, you look at a company like Glean that I know you you're in and is doing well, you know, you can create Glean for the rest of us. If a company commits to being on the Google stack and has all these things and it becomes a combination that no one …”
Gerstner: Google faces the biggest innovator's dilemma in Silicon Valley history
“They are facing the largest innovators dilemma in the history, or certainly from my perspective, the history of Silicon Valley. And I think it's almost impossible to replace a 99% incremental margin business, i.e., a monopoly search business with whatever come…”
Gurley: Gemini has an unmatched advantage in local search reviews
“Gemini is really good at local. Because they have all the local reviews. So I find myself now, if I'm going to a restaurant, you can say, what are the three dishes people have loved the most? What are the two that you stay away from?”
Gerstner: xAI will serve as the hyperscaler powering Musk's companies
“A hundred percent. I think there's no doubt about it. It already is, and it will continue to be. Like, he has a Koretsu, as they say in Japan, of companies that can leverage each other's insights, and I fully expect that they will.”
Gerstner: Superior reasoning architecture beats raw compute scale in AI
“Like you can have the best computers and chips and biggest cluster in the world. But if you don't get the architecture right around your reasoning model and the other guy does, they're probably going to win.”
Gerstner: Nvidia does not intend to compete in frontier AI models
“You know, Nvidia launching a few models. I don't think they intend to be in the, you know, frontier model game where they're competing heads up with these other folks.”
Gerstner: Google and OpenAI will launch superior coding agents in 2025
“I think you're going to see some pretty dramatic breakthroughs this year on coding, right? So think about all the startups that we've seen in the coding space. A lot of them have used prompt injection and other techniques in order to You know, kind of get thes…”