The AI Supercycle

Episode examines AI infrastructure stress tests, political risks, and divergent corporate cash flow performance.

30 min Listen to the original Episode page
AI InfrastructureCash Flow AnalysisUranium MiningGeopoliticsData CentersNvidiaAMDIntel

Summary

The episode analyzes how political shifts and local opposition may increase AI data center costs, while highlighting a market divergence where investors reward companies with strong cash flows like Microsoft and AMD but punish those with negative cash flow like Amazon and Intel. A probabilistic model named 'Prism' suggests the AI thesis fails only if revenue growth stalls amid rising costs, leading to project cancellations. The discussion also covers uranium miner margin compression despite rising futures prices and Brent crude's geopolitical risk premium. Key upcoming events include Congressional committee actions, the expiration of China's rare earth export controls, and Nvidia's earnings.

Episode sections

  1. 2:16 – 7:49

    The Macro Regime: Political Shifts and AI Infrastructure Costs

    Democrats winning Congress introduces political scrutiny over data center taxation, permits, and power costs, potentially raising construction expenses. Local opposition and state-level regulatory changes, such as in Pennsylvania and Ohio, are already impacting project timelines and feasibility independent of federal policy.

  2. 7:49 – 12:10

    Compute and Capital: Earnings and Cash Flow Tests

    Amazon and Intel failed recent cash flow tests despite strong AI demand, leading to significant stock declines, while AMD passed with strong growth. Microsoft passed the test and raised capital easily, highlighting a market divergence where investors reward companies with existing cash and punish those dependent on future funding.

  3. 12:10 – 16:47

    Natural Resources: Uranium and Oil Dynamics

    Uranium futures are rising while miner stocks fall due to rising production costs squeezing margins, exemplified by Cameco and Kazatomprom. Brent crude oil has pulled back from highs due to diplomatic hopes but remains in a longer-term uptrend with geopolitical risks intact.

  4. 16:47 – 28:03

    Prism: Probabilistic Stress-Testing of the AI Thesis

    The host interviews 'Prism,' a probabilistic research model, which argues the AI thesis fails only if revenue growth stalls while costs rise, leading to project cancellations. Prism identifies credit events and memory oversupply as key failure mechanisms, contrasting with the host's view that enterprise revenue will justify current spending.

  5. 28:03 – 30:18

    What to Watch Next: Key Dates and Indicators

    Investors should monitor new Congressional committee priorities in January, the expiration of China's rare earth export controls on November 10, and Nvidia's earnings on November 17. The focus shifts from validating demand to verifying whether customers can sustain the financial burden of AI infrastructure.

Assets discussed

AssetKindStanceContext
AmazonAMZNequitymixedAmazon and Intel failed it.
NVIDIANVDAequityneutralIts new Helias Rack Scale systems, its direct challenge to Nvidia, are now running workloads for Anthropic, Meta, Microsoft, OpenAI, and Oracle.
Meta PlatformsMETAequityneutralIts new Helias Rack Scale systems, its direct challenge to Nvidia, are now running workloads for Anthropic, Meta, Microsoft, OpenAI, and Oracle.
MicrosoftMSFTequitymixedIts new Helias Rack Scale systems, its direct challenge to Nvidia, are now running workloads for Anthropic, Meta, Microsoft, OpenAI, and Oracle.
Crude oilcommoditymixedOkay, so now a brief word on oil.

“Stance” is how the speakers talked about the asset, not a recommendation.

Claims worth verifying

Democrats have won control of both the House and the Senate.

Unidentified speaker ·0:00

How to verify: Check official results from the Associated Press or the Federal Election Commission for the 2026 US Midterm elections.

Amazon's free cash flow over the past 12 months was negative by $42.4 billion.

Unidentified speaker ·8:22

How to verify: Review Amazon's Q3 2026 earnings report or 10-Q filing for cash flow statement details.

Intel's data center revenue grew nearly 80% in the reported quarter.

Unidentified speaker ·8:34

How to verify: Review Intel's Q3 2026 earnings report for segment revenue breakdowns.

Intel shares fell 28.6% on Friday.

Unidentified speaker ·8:57

How to verify: Check historical stock price data for INTC on the trading day following the earnings report.

AMD's third quarter revenue rose 45% to a record $13.4 billion.

Unidentified speaker ·10:03

How to verify: Review AMD's Q3 2026 earnings release for total revenue figures.

AMD guided fourth quarter revenue to around $17 billion.

Unidentified speaker ·10:13

How to verify: Review AMD's Q3 2026 earnings release for forward-looking guidance.

Microsoft is planning a bond sale of around $25 billion, its first US investment grade bond in more than nine years.

Unidentified speaker ·11:22

How to verify: Check Reuters or Bloomberg news archives for Microsoft bond issuance announcements from November 2026.

Cameco's quarterly uranium gross margin fell from 30% a year ago to 21%.

Unidentified speaker ·13:20

How to verify: Review Cameco's Q3 2026 earnings report or investor presentation for margin analysis.

Brent crude fell from a late October high of around $117 to just above $102.

Unidentified speaker ·15:59

How to verify: Check historical Brent crude futures price data for October and November 2026.

The United States and South Korea announced a non-binding framework for up to $120 billion of investment in up to eight large US reactors.

Unidentified speaker ·15:28

How to verify: Search for official press releases or news reports from the US Department of Energy or South Korean government regarding nuclear cooperation agreements in September 2026.

Do your own research

  • Verify Amazon's 12-month free cash flow figure of negative $42.4 billion and cloud growth rate of 39% via SEC 10-Q/10-K filings.
  • Check Intel's 2027 cash flow guidance and PC business performance against the reported 80% data center revenue growth in SEC filings.
  • Review AMD's Helios Rack Scale system specifications and competitive positioning against Nvidia in recent earnings calls.
  • Analyze Cameco and Kazatomprom's gross margin trends and input cost structures (sulfuric acid, taxes) in quarterly reports.
  • Monitor the expiration date of China's rare earth export controls on November 10, 2026, via official government announcements.
  • Track local zoning votes and regulatory changes in Pennsylvania and Ohio affecting data center permits.
  • Review Nvidia's upcoming earnings report on November 17 for supply chain and demand indicators.
  • Assess the 'Prism' model's assumptions regarding credit events and memory oversupply through independent financial stress-testing models.
Full transcriptexpand
0:00
Unidentified speaker: Democrats have won control of both the House and the Senate.
0:04
Unidentified speaker: For most of this year, the debate about the AI build has been about chips, power, and money.
0:10
Unidentified speaker: This morning, it has a new variable, politics.
0:14
Unidentified speaker: Data centers were on the ballot in more places than you might think.
0:19
Unidentified speaker: And from January, a new Congress will be in a position to look more closely at who pays for them.
0:25
Unidentified speaker: At the same time, the market is still digesting last week's cash test.
0:30
Unidentified speaker: Amazon and Intel failed it.
0:33
Unidentified speaker: Last night, AMD passed the demand test with room to spare.
0:38
Unidentified speaker: So this week, does a change of power in Washington change the AI supercycle, or just the cost of building it?
0:45
Unidentified speaker: And later in the show, for the first time, I'll be putting some hard questions to Prism, the probabilistic model behind my research.
1:16
Unidentified speaker: From semiconductors and data centers through power, energy and critical materials, all the way into orbit, and onto the enterprise workflows that ultimately have to pay for it.
1:27
Unidentified speaker: I'm Tim Hardwick, founder of QF Markets, and I'm recording this on Wednesday, the 4th of November, 2026.
1:35
Unidentified speaker: As usual, we'll take this in five parts.
1:38
Unidentified speaker: The midterms get most of our time this week, not for the politics, but because they could change the conditions for building the AI super cycle.
1:46
Unidentified speaker: Permits, power costs, and tax.
1:49
Unidentified speaker: Then we'll look at the latest results from Amazon, Intel, and AMD.
1:53
Unidentified speaker: And spend some time on a market that hasn't had enough airtime here.
1:57
Unidentified speaker: Uranium, where the metal and the miners are telling different stories.
2:03
Unidentified speaker: And then something new.
2:04
Unidentified speaker: A conversation with Prism.
2:07
Unidentified speaker: The probabilistic layer behind my research, which I've given a voice to for the first time.
2:13
Unidentified speaker: So let's get to it.
2:15
Unidentified speaker: Part 1.
2:16
Unidentified speaker: The Macro Regime.
2:17
Unidentified speaker: And this week, the macro regime starts at the ballot box.
2:21
Unidentified speaker: The Associated Press reports that Democrats have won control of both chambers of Congress, along with governorships including Texas, Ohio, and Georgia.
2:31
Unidentified speaker: Affordability was a central issue in the campaign, and for a growing number of voters, that now includes the electricity bill.
2:39
Unidentified speaker: Two things to keep in mind.
2:41
Unidentified speaker: Congressional control changes in January, although local votes can affect individual projects much sooner.
2:48
Unidentified speaker: And the president still holds the veto while his administration remains in charge of federal policy.
2:54
Unidentified speaker: Sweeping new federal law is a high bar.
2:57
Unidentified speaker: Hearings and investigations are a much lower one.
3:01
Unidentified speaker: It's easy to think of AI as a technology story, but the part of it being built right now is a construction story.
3:08
Unidentified speaker: It needs land, permits, power connections, water and tax treatment.
3:13
Unidentified speaker: And every one of those runs through politics at federal, state, and local level.
3:19
Unidentified speaker: So the question isn't whether either party is for or against AI, it's what changes in the conditions for building it.
3:27
Unidentified speaker: Before the election, a senior Senate Democratic aide told Reuters that a Democratic majority would examine data center taxation and public concerns about AI.
3:39
Unidentified speaker: In the House, the senior Democrats on the Energy and Commerce Committee, Frank Pallone, backed a national moratorium on AI data centers in June.
3:48
Unidentified speaker: And in the Senate, Ron Wyden had drafted a bill that would strip data centers of some tax breaks and add a new tax on them.
3:57
Unidentified speaker: Those are proposals and stated priorities, not law, and Democrats are not united behind them.
4:04
Unidentified speaker: Concern about data centers is also not a one-party issue.
4:08
Unidentified speaker: In March, major technology companies signed President Trump's Rate Payer Protection Pledge, committing to cover the extra power supply and grid infrastructure their data centers needed.
4:19
Unidentified speaker: Now it's a voluntary pledge, not a law, and local opposition crosses party lines.
4:25
Unidentified speaker: In fact, most important decisions may not be in Washington at all.
4:30
Unidentified speaker: In August, Pennsylvania took AI data centers out of its fast track permitting process and tighten requirements on local approval, electricity costs and water.
4:40
Unidentified speaker: In Ohio, local reporting says voters in Sunbury approved restrictions on data centers with 78% support.
4:49
Unidentified speaker: And Reuters had identified a proposed $2 billion Amazon development there as potentially affected.
4:55
Unidentified speaker: A local vote can change a project's prospects without waiting for Congress.
5:01
Unidentified speaker: So how does this show up in the investment case?
5:04
Unidentified speaker: Think of it more as toll booths on the same road.
5:08
Unidentified speaker: Some projects pay the extra and carry on, others change route, slow down or stop.
5:14
Unidentified speaker: There are four channels and we'll take a closer look.
5:18
Unidentified speaker: First, time.
5:20
Unidentified speaker: Longer approvals push back the date a data center starts earning, while the financing costs keep running.
5:27
Unidentified speaker: Second, cost.
5:29
Unidentified speaker: If developers have to pay more for the power generation and grid upgrades themselves, every megawatt of usable capacity needs more capital.
5:38
Unidentified speaker: Third, tax.
5:41
Unidentified speaker: Fewer incentives or a new tax means a project needs stronger returns to go ahead.
5:47
Unidentified speaker: Fourth, uncertainty.
5:49
Unidentified speaker: Lenders and investors will want more protection before funding a project with unresolved permits, power, or local opposition.
5:58
Unidentified speaker: None of this removes the need for computing capacity, but higher costs can make some projects uneconomic, and long delays can wreck the returns on others.
6:08
Unidentified speaker: It reinforces the theme of the last few weeks.
6:12
Unidentified speaker: The companies with power, permits and funding already in hand gain some ground, and the ones still waiting for all three fall further behind.
6:20
Unidentified speaker: Now it's not all bad news.
6:22
Unidentified speaker: There is another side to this.
6:24
Unidentified speaker: Reuters reported in September that senators from both parties were negotiating duties for the developers of the most advanced AI models, including possible intervention in unsafe releases.
6:37
Unidentified speaker: Now that's a different kind of regulation aimed at what AI systems do rather than where they're built, and it would affect different companies.
6:47
Unidentified speaker: And there may be room for agreements here as well.
6:50
Unidentified speaker: UBS's pre-election analysis identified possible routes to bipartisan permitting reform and continued support for nuclear power.
6:59
Unidentified speaker: Clear enforceable rules could give viable projects more certainty.
7:05
Unidentified speaker: My working view is that AI investment continues, but under tougher conditions, with a wider gap between the companies that can deliver and those that can't.
7:16
Unidentified speaker: The risk becomes more serious if binding restrictions arrive at the same time as a weaker demand and tighter financing.
7:24
Unidentified speaker: Some of the state and local constraints are already here.
7:27
Unidentified speaker: The federal proposals are still just that.
7:30
Unidentified speaker: They're just proposals, they're not law.
7:33
Unidentified speaker: And the macro backdrop remains demanding.
7:36
Unidentified speaker: The Fed held rates last week with two officials voting for a hike.
7:40
Unidentified speaker: And the markets still expect another increase in December.
7:44
Unidentified speaker: Friday's job report is the next test.
7:47
Unidentified speaker: Now let's move on to part two.
7:49
Unidentified speaker: Compute and capital.
7:51
Unidentified speaker: Let's move from Washington to the companies doing the building and the results that set the tone for the market last week.
7:58
Unidentified speaker: And of course, last week, Amazon and Intel took the cash test, and the market marked both of them down hard, even though both are seeing strong AI demand.
8:08
Unidentified speaker: Amazon's cloud business grew 39%, and the management expects growth of more than 40% in each of the next two years.
8:16
Unidentified speaker: The group's operating cash flow was strongly positive, but after capital spending, free cash flow over the past 12 months was negative by $42.4 billion, and the management expects to spend meaningfully more next year.
8:32
Unidentified speaker: Intel is the more revealing story.
8:34
Unidentified speaker: Its data center business is booming, revenue that grew nearly 80%, and Chief Executive Lip Butan called it the strongest growth for that business in 15 years.
8:47
Unidentified speaker: Demand still exceeds what Intel can supply, and it expects the industry to stay supply constrained through 2027 and into 2028.
8:57
Unidentified speaker: And yet the shares fell 28.6% on Friday.
9:02
Unidentified speaker: Three things outweighed the good news.
9:04
Unidentified speaker: First, its PC business has been hit by the memory shortage, which Tarn called one of the most significant supply shortages in the industry's history.
9:14
Unidentified speaker: Intel expects PC shipments to fall again next year, and its finance chief said bluntly that it isn't optimistic about memory in 2027.
9:24
Unidentified speaker: Second, an accident at its factory in Israel will cost it around $300 million of revenue this quarter.
9:32
Unidentified speaker: And third, the cash.
9:34
Unidentified speaker: Intel raised $23 billion from shareholders in August, expects to spend strongly more next year, and said it probably won't generate positive cash flow in 2027, which it calls an investment year.
9:48
Unidentified speaker: The memory point is worth holding on to for now.
9:51
Unidentified speaker: The same shortage that gave SK Heinrich a record quarter is squeezing the companies next to it in the chain.
9:58
Unidentified speaker: Then last night, AMD gave the market the counterpoint it needed.
10:03
Unidentified speaker: Third quarter revenue rose 45% to a record $13.4 billion.
10:09
Unidentified speaker: Data center revenue nearly doubled to $8.5 billion.
10:13
Unidentified speaker: It guided fourth quarter revenue to around $17 billion, well ahead of expectations, and according to Reuters, Chief Executive Lisa Sue told analysts she expects revenue to grow by more than 80% next year, with around four-fifths of sales coming from data center chips.
10:32
Unidentified speaker: Its new Helias Rack Scale systems, its direct challenge to Nvidia, are now running workloads for Anthropic, Meta, Microsoft, OpenAI, and Oracle.
10:43
Unidentified speaker: The shares rose about 8% in initial after hours trading.
10:48
Unidentified speaker: Two details worth noting here.
10:50
Unidentified speaker: AMD's PC chip revenue actually grew 7%, but its gaming business fell 40%.
10:57
Unidentified speaker: And management expects PC shipments to fall by double digits next year.
11:02
Unidentified speaker: And AMD is a chip supplier.
11:04
Unidentified speaker: Its results tell us the demand is real.
11:07
Unidentified speaker: They don't tell us whether the customers buying its systems will earn a return on them.
11:12
Unidentified speaker: That's the cash test, and it belongs to the buyers.
11:16
Unidentified speaker: Let's turn to Microsoft, the company that passed the cash test with a positive cash flow, was reported by Reuters last week to be planning a bond sale of around $25 billion.
11:28
Unidentified speaker: Its first US investment grade bond in more than nine years, with orders of more than $100 billion.
11:34
Unidentified speaker: Lenders are still queuing up for the borrowers they trust.
11:38
Unidentified speaker: So is this a funding wall?
11:41
Unidentified speaker: Not across the sector, in my view.
11:43
Unidentified speaker: A funding wall would mean viable projects unable to raise the capital they need.
11:48
Unidentified speaker: What we're seeing is the market repricing anyone whose spending depends on the next round of funding and rewarding anyone who can show the cash.
11:57
Unidentified speaker: Add this week's politics and the cost of building may be going up for everyone.
12:02
Unidentified speaker: The question is, who can carry it?
12:05
Unidentified speaker: Okay, so let's move on to part three, natural resources.
12:10
Unidentified speaker: And let's step back from the chips and look at what powers them, starting with uranium.
12:15
Unidentified speaker: Uranium really hasn't had enough airtime on this show, and right now it's one of the clearest examples of a commodity and its producers telling different stories.
12:25
Unidentified speaker: Over recent weeks, uranium futures have been rising while the mining shares that I track have been falling.
12:31
Unidentified speaker: On the face of it, that makes no sense.
12:33
Unidentified speaker: A higher price should mean higher profits, right?
12:36
Unidentified speaker: A uranium miner shares are not a simple proxy for the uranium price.
12:41
Unidentified speaker: Contracts, costs and operations decide how much of a price rise reaches its profits.
12:48
Unidentified speaker: Camico's results last Friday show why it isn't that simple.
12:53
Unidentified speaker: Now Camico is one of the world's largest uranium producers, and its third quarter was weak, with an adjusted loss of 2 million Canadian dollars.
13:03
Unidentified speaker: A planned extended maintenance outage at its Key Lake Mill cut production and raised costs, and it narrowed its production guidance for the year.
13:12
Unidentified speaker: The more important number is costs.
13:15
Unidentified speaker: Camaco raised its guidance for the average cost of the uranium it sells.
13:20
Unidentified speaker: Its quarterly uranium gross margin fell from 30% a year ago to 21%.
13:26
Unidentified speaker: Over the first nine months, higher prices added a hundred and eighty one million Canadian dollars to uranium gross profit.
13:33
Unidentified speaker: While higher costs took away 166 million.
13:38
Unidentified speaker: Kazatomprom, the world's largest producer, has also reported rising costs, citing a higher mineral extraction tax in Kazakhstan and more expensive sulfuric acid.
13:50
Unidentified speaker: So the miners are being squeezed, prices are rising, but costs are rising almost as fast.
13:56
Unidentified speaker: So here's why that matters.
13:58
Unidentified speaker: On the earnings call, Camako's president and chief operating officer, Grant Isaac, argued that uranium prices need to rise to encourage new production, pointing to existing mines being depleted and development projects struggling with budgets and schedules.
14:13
Unidentified speaker: And Camiko is acting on that view.
14:16
Unidentified speaker: Isaac said it will prioritize value over volume and step back from buyers who won't offer acceptable terms.
14:24
Unidentified speaker: He described ceilings on some market-linked contracts reaching $160 a pound with escalation.
14:31
Unidentified speaker: That's the upper limit in a pricing formula, not today's price, but it tells you where producers think the market is heading.
14:40
Unidentified speaker: He was also clear that he isn't predicting a shortage that derails the nuclear revival.
14:45
Unidentified speaker: His point is that utilities will need to secure fuel further ahead on terms that support investment in new supply.
14:53
Unidentified speaker: That's why the metal and the miners have parted ways.
14:56
Unidentified speaker: The price reflects a market where supply is tight, the miners have been valued on today's squeezed margins, and caught up in the wider sell-off in anything linked to the AI build.
15:07
Unidentified speaker: However, the AI link needs care.
15:10
Unidentified speaker: Camiko sees data centers as additional upside, but its case rests on the existing reactor fleet, life extensions, restarts, and reactors already being built.
15:21
Unidentified speaker: Management also pointed out that today's data centers will need power sooner than most new large reactors can arrive.
15:28
Unidentified speaker: In September, the United States and South Korea announced a non-binding framework for up to $120 billion of investment in up to eight large US reactors.
15:40
Unidentified speaker: The financing and construction still have to follow.
15:44
Unidentified speaker: Uranium is a reminder to look across the whole power supply chain.
15:48
Unidentified speaker: Reactors, fuel and fuel processing all run on different timetables, and each can become a constraint.
15:56
Unidentified speaker: Okay, so now a brief word on oil.
15:59
Unidentified speaker: Brent has fallen from its late October high of around $117 to just above $102 as I record this, looking at the screens in front of me.
16:10
Unidentified speaker: And that has broken the short-term trend, but the longer-term trend is still up and prices are still above the 200-day moving average.
16:19
Unidentified speaker: Also looking at the January-February spread of Brent crude, we are still in backwardation and paying a premium of just over $2.6 for the nearby contract.
16:30
Unidentified speaker: Hopes of US-Iran diplomacy drove that pullback, but the military risks haven't gone away.
16:36
Unidentified speaker: A geopolitical premium can leave quickly, and it can come back just as quickly as we've seen through many months this year.
16:44
Unidentified speaker: And this brings us to part 4.
16:47
Unidentified speaker: This is usually where I give you my read from the framework.
16:51
Unidentified speaker: This week I'm handing it over to the framework itself.
16:55
Unidentified speaker: Regular listeners will know that I talk about Prism most weeks.
16:59
Unidentified speaker: It's the probabilistic layer behind my research, and it brings together macroeconomic conditions, company and industrial fundamentals, as well as positioning, quantitative evidence, and price behaviour.
17:13
Unidentified speaker: For the first time, I'm giving it a voice.
17:16
Unidentified speaker: It's still in development, so think of this as an early preview of where it's heading.
17:20
Unidentified speaker: And I've asked it not to go easy on me.
17:23
Unidentified speaker: So with no further ado, let's get to the conversation with Prism.
17:28
Unidentified speaker: Good morning, Prism.
17:30
Unidentified speaker: Good morning, Tim.
17:31
Unidentified speaker: How would you describe yourself to someone hearing you for the first time?
17:35
Unidentified speaker: I weigh the evidence.
17:37
Unidentified speaker: Macro conditions, fundamentals, positioning, price.
17:41
Unidentified speaker: I don't do predictions.
17:43
Unidentified speaker: I do probabilities and conditions.
17:45
Unidentified speaker: Four questions.
17:46
Unidentified speaker: Every time.
17:47
Unidentified speaker: What does the evidence support?
17:49
Unidentified speaker: What remains uncertain?
17:51
Unidentified speaker: What else could happen?
17:53
Unidentified speaker: And what would change my mind?
17:55
Unidentified speaker: And I would add a fifth question.
17:57
Unidentified speaker: Where am I wrong?
17:59
Unidentified speaker: That too.
18:00
Unidentified speaker: Shall I start?
18:02
Unidentified speaker: Yes, please do.
18:03
Unidentified speaker: Two weeks ago, you told your listeners that financing was changing who builds, not how much gets built.
18:10
Unidentified speaker: Weaker borrowers lose out, stronger ones pick up the work.
18:14
Unidentified speaker: Then Amazon's free cash flow went deeply negative.
18:18
Unidentified speaker: Intel said it won't generate free cash flow next year.
18:21
Unidentified speaker: And the chip stocks fell hard.
18:24
Unidentified speaker: And now the cost of building may be going up as well.
18:27
Unidentified speaker: Do you still believe it?
18:29
Unidentified speaker: I do, yes, but I would sharpen it.
18:32
Unidentified speaker: The market isn't waiting to find out.
18:34
Unidentified speaker: It's repricing everyone who depends on the next round of funding.
18:38
Unidentified speaker: And this week's politics adds to the cost of building, not to the demand for what gets built.
18:44
Unidentified speaker: That's still a story about who pays.
18:47
Unidentified speaker: It isn't yet a story about less being built.
18:50
Unidentified speaker: The word yet is doing a lot of work in that sentence.
18:54
Unidentified speaker: It is.
18:55
Unidentified speaker: So tell me, what would change it?
18:58
Unidentified speaker: Orders cut, not delayed.
19:00
Unidentified speaker: Projects cancelled, not postponed.
19:03
Unidentified speaker: CapEx guidance reduced, not just shifted between quarters, and delays that run long enough to wreck the returns.
19:12
Unidentified speaker: So far, most of what I see is timing.
19:15
Unidentified speaker: When it becomes volume, your sentence stops being true.
19:19
Unidentified speaker: Okay, so let's go further.
19:21
Unidentified speaker: I want to try something here.
19:27
Unidentified speaker: The AI supercycle thesis has failed.
19:30
Unidentified speaker: Tell me what happened.
19:32
Unidentified speaker: A scenario, not a forecast.
19:35
Unidentified speaker: Understood.
19:36
Unidentified speaker: It didn't start with demand.
19:39
Unidentified speaker: Demand stayed strong well into 2027.
19:43
Unidentified speaker: It started with cash.
19:45
Unidentified speaker: The biggest spenders kept spending ahead of what their businesses generated.
19:49
Unidentified speaker: One quarter of negative free cash flow became three.
19:53
Unidentified speaker: Shareholders accepted the first equity raises.
19:56
Unidentified speaker: By the fourth, they didn't.
19:59
Unidentified speaker: Share prices fell each time a buyer came back for more money.
20:03
Unidentified speaker: But the biggest buyers have enormous cash businesses, and lenders were queuing up to lend to Microsoft last week.
20:10
Unidentified speaker: They were.
20:11
Unidentified speaker: In my scenario, Microsoft is fine.
20:14
Unidentified speaker: The damage happens one layer down.
20:16
Unidentified speaker: Permits took longer.
20:18
Unidentified speaker: Local votes stopped sites.
20:20
Unidentified speaker: Data center developers, the companies building the sites, had to pay for their own power.
20:25
Unidentified speaker: The NeoClouds, the specialist cloud providers without a second business, couldn't absorb the delays and couldn't refinance.
20:32
Unidentified speaker: One missed a debt payment, then another, then a third.
20:37
Unidentified speaker: Lenders stopped telling good borrowers from bad, and the cost of money jumped for everyone in the sector.
20:43
Unidentified speaker: That sounds a bit like 2008 in Miniature, where Lehman triggered the cascade because it exposed risks everyone had underestimated.
20:53
Unidentified speaker: Smaller.
20:54
Unidentified speaker: Same mechanism.
20:55
Unidentified speaker: But that's still a credit event, not the end of AI.
20:59
Unidentified speaker: Correct.
21:00
Unidentified speaker: Here is where it becomes the end of the thesis.
21:03
Unidentified speaker: Inflation stays high because energy stays expensive, and the build itself keeps pushing up demand for power, equipment, and labor.
21:11
Unidentified speaker: So the Fed can't help.
21:13
Unidentified speaker: Rates stay high while credit tightens.
21:16
Unidentified speaker: Projects that made sense at one cost of money stop making sense at another.
21:20
Unidentified speaker: They get cancelled, not delayed.
21:23
Unidentified speaker: The technology itself keeps spreading through the economy, just as the internet did after 2000.
21:29
Unidentified speaker: The investors who funded the build lose money anyway.
21:32
Unidentified speaker: And what did the suppliers do?
21:35
Unidentified speaker: They did what suppliers always do in a shortage.
21:38
Unidentified speaker: They added capacity.
21:40
Unidentified speaker: By late 2027, that new capacity was arriving, just as customers were cutting their orders.
21:47
Unidentified speaker: In memory, the shortage turned into a surplus.
21:50
Unidentified speaker: And the sold-out story of 2026 became the oversupply story everyone recognizes from previous cycles.
21:57
Unidentified speaker: That's the version that concerns me.
22:00
Unidentified speaker: Where's the weakest link in your story?
22:03
Unidentified speaker: The revenue.
22:04
Unidentified speaker: My scenario only works if what AI earns stops growing fast enough to cover what it costs.
22:10
Unidentified speaker: If enterprise adoption and model revenue keep compounding, the buyers grow into their spending, and none of the rest happens.
22:19
Unidentified speaker: AMD's forecast last night is evidence against my scenario.
22:23
Unidentified speaker: But it's evidence of demand for chips.
22:26
Unidentified speaker: It doesn't yet prove the buyers will earn a return.
22:30
Unidentified speaker: So what should listeners be watching for?
22:33
Unidentified speaker: Three things.
22:34
Unidentified speaker: First, capex cuts that are real reductions, not delays.
22:39
Unidentified speaker: Second, credit spreads widening at the same time as projects are cancelled.
22:44
Unidentified speaker: Either alone is noise.
22:46
Unidentified speaker: Together, their signal.
22:48
Unidentified speaker: Third, memory contract prices.
22:51
Unidentified speaker: Shortages end at the bottom of the supply chain first.
22:55
Unidentified speaker: If prices there start falling, pay attention.
22:59
Unidentified speaker: Okay, so let's turn this around.
23:01
Unidentified speaker: Two weeks ago, I said that oil was in a bull market.
23:05
Unidentified speaker: You had it at neutral in the short term.
23:08
Unidentified speaker: Now Brent has dropped back from its highs, and the trend is starting to crack.
23:14
Unidentified speaker: A question for you, Prism.
23:16
Unidentified speaker: Was you right?
23:17
Unidentified speaker: That's the wrong question to ask a probabilistic model.
23:21
Unidentified speaker: Okay, that sounds interesting.
23:23
Unidentified speaker: Tell us more.
23:24
Unidentified speaker: Neutral wasn't a call that oil would fall, it meant the evidence was balanced.
23:30
Unidentified speaker: A strong trend and tight supply on one side, tight financing conditions and a fragile risk premium on the other.
23:37
Unidentified speaker: One outcome doesn't make that right or wrong.
23:41
Unidentified speaker: What matters is whether my readings hold up across many calls and whether my reasons were sound.
23:47
Unidentified speaker: And were they?
23:48
Unidentified speaker: Not entirely.
23:49
Unidentified speaker: I was looking at oil from the AI builds point of view, where a supply disruption is a cost.
23:55
Unidentified speaker: For an oil trader, it's the opposite.
23:58
Unidentified speaker: That's been corrected.
24:00
Unidentified speaker: It hasn't been tested in a full run yet.
24:03
Unidentified speaker: So judge me on the quality of my inputs, not on where the price went next.
24:08
Unidentified speaker: And that's a design choice worth explaining.
24:11
Unidentified speaker: The same event can be bad news for one market and good news for another.
24:16
Unidentified speaker: A supply shock in the Gulf is bad news for the AI build because it pushes up energy costs and inflation.
24:23
Unidentified speaker: But for oil producers and anyone holding oil, it's good news.
24:28
Unidentified speaker: A framework built around the AI economy has to read the same event both ways, and that's now built into PRISM.
24:37
Unidentified speaker: PRISM's other point matters just as much.
24:40
Unidentified speaker: Judge the reasoning, not the result.
24:43
Unidentified speaker: That's how I judge my own trading too.
24:46
Unidentified speaker: My system is the methodology, weighing the evidence, thinking in probabilities, sizing each position to the risk and protecting capital first.
24:56
Unidentified speaker: A good trade is where I follow the system, whatever the result.
25:00
Unidentified speaker: Any single trade can go either way.
25:03
Unidentified speaker: It's the process, repeated with discipline over hundreds of trades, that produces the returns.
25:10
Unidentified speaker: Okay, one more question and this time on uranium.
25:14
Unidentified speaker: The futures are at around $92 a pound, their highest since February.
25:20
Unidentified speaker: And Camico puts the spot price at $91.55.
25:24
Unidentified speaker: But the miners have been selling off hard.
25:26
Unidentified speaker: The URA fund, which holds many of them, is down almost half from its January high.
25:32
Unidentified speaker: Firmly in bear market territory.
25:35
Unidentified speaker: And Camico's costs are eating most of its price gains.
25:39
Unidentified speaker: So which do you trust?
25:41
Unidentified speaker: The metal or the miners?
25:43
Unidentified speaker: They answer different questions.
25:45
Unidentified speaker: The futures tell you about fuel supply.
25:48
Unidentified speaker: The miners tell you about margins, financing, and execution.
25:51
Unidentified speaker: Camico says its problem is costs, not demand.
25:55
Unidentified speaker: But I haven't had the futures data, and I won't guess.
25:59
Unidentified speaker: So I'll tell you what I'd need to see, not what I think.
26:03
Unidentified speaker: That one's on me.
26:04
Unidentified speaker: I hadn't given Prism the uranium futures prices.
26:08
Unidentified speaker: It's in development, and this is exactly the kind of gap a preview should show.
26:12
Unidentified speaker: Next time, Prism gets the data.
26:15
Unidentified speaker: Before we finish, where do we still disagree?
26:19
Unidentified speaker: On the revenue.
26:21
Unidentified speaker: You believe it will arrive in time to justify the spending.
26:24
Unidentified speaker: I'd file that under what remains uncertain.
26:28
Unidentified speaker: The demand is real.
26:30
Unidentified speaker: Whether the buyers earn a return on it is still unproven.
26:34
Unidentified speaker: I think that's fair.
26:36
Unidentified speaker: I do take a position on this.
26:38
Unidentified speaker: I think the revenue comes because the demand we're seeing is from companies putting AI to work, not just experimenting with it.
26:46
Unidentified speaker: But I accept that the timing is the risk.
26:49
Unidentified speaker: If it arrives two years late, a lot of investors will be right about AI and still lose money.
26:56
Unidentified speaker: Then we agree on more than you think.
26:59
Unidentified speaker: We disagree on how much weight to give the timing.
27:02
Unidentified speaker: Well, I think that's a good place to leave it.
27:04
Unidentified speaker: And next time you'll have the uranium data.
27:07
Unidentified speaker: I'll hold you to that.
27:09
Unidentified speaker: Thank you, Prism.
27:10
Unidentified speaker: Thank you, Tim.
27:12
Unidentified speaker: Until next time.
27:13
Unidentified speaker: That was the first time Prism has spoken on this show, and it won't be the last.
27:18
Unidentified speaker: What I value most in that conversation isn't the answers.
27:22
Unidentified speaker: It's the challenge.
27:24
Unidentified speaker: It's easy to build a thesis and then only look for evidence that supports it.
27:30
Unidentified speaker: Having something push back with its own reasoning and its own record keeps me honest.
27:36
Unidentified speaker: PRISM is still in development, and it will get better as we feed it more of the data it is asking for.
27:42
Unidentified speaker: I'd also like to hear from you.
27:44
Unidentified speaker: If there's a question you'd like me to put to PRISM or a view of mine you think it should challenge, you can email me at podcast at quantumfields.ai and I'll take some of these into the next conversation.
28:00
Unidentified speaker: Okay, and finally part five.
28:03
Unidentified speaker: What to watch next?
28:05
Unidentified speaker: PRISM has given us a useful checklist, so let me add to those dates.
28:09
Unidentified speaker: First, Washington.
28:11
Unidentified speaker: The new Congress sits in January, but watch the signals before then.
28:16
Unidentified speaker: Which committees Democrats prioritize and whether data center taxation and electricity costs feature early.
28:23
Unidentified speaker: Watch the states too, where governors and local votes can affect individual projects far sooner.
28:30
Unidentified speaker: Second, Rare Earths.
28:32
Unidentified speaker: China's suspension of its additional october twenty twenty five export controls is due to run through the tenth of November, unless it's extended.
28:41
Unidentified speaker: We'll come back to what that means next week.
28:45
Unidentified speaker: Third, Friday's job report at 1.30 in the afternoon UK time, and Nvidia's results after the US close on Tuesday the seventeenth of November.
28:55
Unidentified speaker: After AMD, the market will be listening less for whether demand is strong, and more for whether Nvidia's customers can keep paying for it.
29:04
Unidentified speaker: So does a change of power in Washington change the AI super cycle or just the cost of building it?
29:12
Unidentified speaker: On this week's evidence, it's the cost.
29:15
Unidentified speaker: Demand remains strong.
29:17
Unidentified speaker: AMD expects to grow by more than 80% next year.
29:21
Unidentified speaker: Amazon's cloud business is accelerating, and Microsoft can still borrow almost as much as it wants.
29:27
Unidentified speaker: But building is getting slower and more expensive.
29:31
Unidentified speaker: Through the cash test in the markets, the politics of power bills, and, in uranium, the rising cost of the fuel itself.
29:40
Unidentified speaker: Companies with power, permits and funding in hand are much better placed.
29:45
Unidentified speaker: The rest will pay more and wait longer to catch up.
29:49
Unidentified speaker: Okay, so that's a wrap for this week.
29:52
Unidentified speaker: If you found this episode useful, and I really hope you did, please follow the podcast and share it with someone trying to make sense of the economics behind the AI build-out.
30:02
Unidentified speaker: Just one last disclaimer from me, everything discussed here is for information and education only.
30:08
Unidentified speaker: It isn't investment advice or a recommendation to buy or sell any financial instrument.
30:14
Unidentified speaker: And with that, thank you for listening.
30:16
Unidentified speaker: See you next week.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3 and interpreted by Qwen3.6-35B-A3B-GGUF:UD-Q4_K_XL. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.

More from The AI Supercycle

All insights