Uranium: the fuel behind the AI investment boom
Chris Frostad on the GG Podcast: tech-driven nuclear demand meets a structural supply gap, the uranium investment ladder from physical trusts to explorers, why Purepoint partners with majors, what makes the Athabasca Basin different, and a five-year study of how the market actually digests junior press releases.
One-line take: Frostad's case is that nuclear has moved from talk to action — tech companies are contracting power (Microsoft and Three Mile Island, Amazon, Google), SMRs and micro-reactors are arriving, governments are extending reactor lives — and that demand is highly predictable while supply is not: the world has produced 20–30% less uranium than it burned for 3–4 years, living off post-Fukushima inventories, so "we are right now in the middle of a structural supply gap" that takes a decade to close because a new mine takes ten years from discovery. Explorers carry the most torque ("hundreds to thousands times payback" — NextGen went from $40M to ~$10B) but must be judged on capital structure, number of projects, management and partners. Purepoint's own pitch: every primary project is joint-ventured with a major (Cameco, Orano, IsoEnergy) with Purepoint as paid operator, which cuts dilution and doubles as third-party validation. The most useful nugget is his AI study of 650 press releases from 40 Basin juniors over five years: hype moves the stock on day one, "by day three it's all lost, and by day five it's actually where it's supposed to be." This is the CEO of a junior explorer talking his own book; host Gary Gill discloses he holds spot uranium, explorers and Purepoint shares. (Back-filled 2026-SEP-18 — the episode predates the hub's first Spotlight page.)
1. Stocks & names mentioned
Unlike the scripted Uranium Spotlight briefs, this is an interview in which Frostad pitches his own company, so Purepoint (PUR.V) is tabled here as the management view. The producers and developers are named as rungs on his "investment ladder," not as recommendations; the tech names are cited as demand drivers.
| Ticker | Name | Research | View | What he said | At |
|---|
| PUR.V | Purepoint Uranium Group (PUR: TSXV; PTUUF: OTC) | SA · STK | Positive | His own company (CEO; management view). Every primary project is joint-ventured with a major and Purepoint operates: a ~100,000 ha 50/50 district with IsoEnergy next to the 45% Hurricane deposit (new Nova zone, up to 8% U), Smart Lake 27% with Cameco, Hook Lake 21% with Cameco and Orano. On a $3M program it pays $1.5M and gets a 10% operator fee back, so net cost is $1.2M — "I'd rather own 27% of a deposit than 100% of some blank real estate." | 15:14 |
| Uranium | Uranium (U3O8 — commodity) | — | Positive | For 3–4 years the world produced 20–30% less than it consumed, covered by post-Fukushima inventories and cheap Kazakh supply that pushed price below production cost; as inventories run out, "we are right now in the middle of a structural supply gap." Demand is predictable (a reactor is "a customer for life"), but a new mine takes a decade, so the gap "can't be repaired in the near term… you're going to see the price of uranium go through the roof." | 03:52 |
| CCJ | Cameco | QT · SA · STK · FA | Positive | First name on the producer rung — producers are "the most immediately impacted by the price of uranium rising." Also Purepoint's JV partner at Smart Lake (73%) and Hook Lake. | 08:41 |
| KAP | Kazatomprom (KAP: LSE GDR) | SA · STK | Positive | Named on the producer rung ("which trades in London"), directly levered to a rising uranium price; Kazakhstan supplies ~40% of world uranium, and its past flood of low-cost pounds helped push price below production cost. | 08:41 |
| UUUU | Energy Fuels | QT · SA · STK · FA | Positive | A US producer on the producer rung — one of "no more than a small handful" of producers, "directly impacted by the price of uranium." | 08:41 |
| UEC | Uranium Energy Corp | QT · SA · STK · FA | Positive | Named alongside Cameco, Kazatomprom and Energy Fuels as a producer — the rung most immediately lifted by a rising uranium price. | 08:41 |
| DNN | Denison Mines | QT · SA · STK · FA | Positive | Developer rung: companies with defined resources turning them into mines — "the ones that we're relying on to fill a lot of that gap." Only a handful exist. | 09:06 |
| NXE | NextGen Energy | QT · SA · STK · FA | Positive | A developer, and his proof of explorer upside: "NextGen, which was a $40 million company, is now worth 10 billion dollars… over a matter of a short period of time." | 11:37 |
| ISOU | IsoEnergy (ISO: TSX) | QT · SA · STK · FA | Positive | Found the Basin's latest discovery, Hurricane: a little under 50M lb at a 45% average grade — "a bit of a freak show." Purepoint's 50/50 partner in the ~100,000 ha district beside it. | 10:07 |
| SRUUF | Sprott Physical Uranium Trust (SPUT — U.UN/U.U: TSX) | SA · STK | Neutral | Top rung of his ladder: you can't keep uranium "under our mattress," but SPUT puts all the money it raises into buying and storing physical uranium — "pretty much a pure investment in the commodity itself and it's one of the few ways you can do that." | 07:26 |
| Orano | Orano (French state-owned — private) | — | Neutral | "Huge nuclear French company," co-owner with Cameco of the Hook Lake JV where Purepoint holds 21% and operates; majors like Orano "wouldn't be writing checks for projects that they didn't think had the potential to deliver a tier one asset." | 15:55 |
| MSFT | Microsoft | QT · SA · STK · FA | Neutral | Demand-side example: Three Mile Island had shut because it wasn't economic, but when Microsoft contracted its power for years "all of a sudden it became economic to do that again." | 00:37 |
| AMZN | Amazon | QT · SA · STK · FA | Neutral | "The Amazons of the world, the Googles of the world" are "getting right into" nuclear because data centers need immense power that "doesn't exist right now" — "you can't just plug it into the grid. The lights would all dim." | 00:58 |
| GOOGL | Alphabet (Google) | QT · SA · STK · FA | Neutral | Named with Amazon as a hyperscaler moving directly into nuclear power to feed data centers — the first choke point before fuel supply becomes the next one. | 00:58 |
2. Talking points
00:17 More action than talk on nuclear
- A few years ago the "nuclear renaissance" was mostly talk about the environment and base load; the difference now is that tech is in it.
- Microsoft's contract made Three Mile Island economic to restart 00:37; Amazon and Google are "getting right into it" because data centers need power that doesn't exist on the grid.
01:41 SMRs, micro-reactors and life extensions
- Small modular reactors deploy quicker and cheaper; micro-reactors can be "drag[ged] behind a truck" to power remote mines or communities — one is slated for the moon.
- Governments are backing nuclear "who weren't before" and extending reactors that were due to shut 02:23 — and all of it comes back to one question: where does the fuel come from?
03:07 Demand is predictable; supply isn't
- After power, the next choke point down the chain is fuel. A reactor is "a customer for life or at least for 40 years," and its annual uranium needs are known in advance 03:31.
- For 3–4 years the world produced 20–30% less than it consumed 03:52, bridged by inventories built after Fukushima (Japan shut ~10% of the world fleet) while Kazakhstan flooded the market — pushing price below production cost.
04:33 A structural supply gap with a ten-year fuse
- As inventories run out, "we are right now in the middle of a structural supply gap." Unlike most markets, higher prices can't quickly pull in supply — "it takes a decade to turn a mine on from nothing" 04:53.
- Money must first find the uranium (years), then develop it (more years) 05:15; the deficit "can't be repaired in the near term… maybe over the next 10, 12 years" and price will "go through the roof" 05:59. Purepoint's "Uranium Spotlight" site carries primers and white papers on this.
07:26 The investment ladder, top-down
- Physical: Sprott Physical Uranium Trust — "pretty much a pure investment in the commodity." Then broad-metal, nuclear, junior-mining and nuclear-utility ETFs 07:56.
- Producers 08:41 (Cameco, Kazatomprom, Energy Fuels, Uranium Energy) — "no more than a small handful," most immediately hit by price. Developers 09:06 (Denison, NextGen) fill the gap. Explorers carry "the most torque," potentially "100, a thousandfold."
09:35 Why the explorers cluster in the Athabasca Basin
- Highest grades in the world and still the most discoveries: ~40 deposits in 50 years totaling ~2 billion lb, "and they're not getting smaller."
- Latest: IsoEnergy's Hurricane 10:07 — a little under 50M lb at a 45% average grade.
10:52 Judge explorers through a different lens
- Explorers don't rise with the commodity tide — "there's nothing in the warehouse yet." Look for a tight capital structure (constant share issuance means "you run out of highway"), many projects not one, management, and how and how often results are presented 11:16.
- Millions are spent and dozens of holes drilled before the first good hole. NextGen went from $40M to ~$10B 11:37. Don't put all your eggs in one explorer; if one hits, follow it closely to make sure it isn't a fluke 13:18.
12:15 Purepoint's model: partner with majors and operate
- Primary projects are joint-ventured with Cameco, Orano and IsoEnergy; Purepoint is the operator — "their exploration arm" — with the majors "watching over our shoulder."
- The IsoEnergy district is ~100,000 ha, 50/50, home to the new Nova zone (up to 8% U) 15:14. On a $3M program Purepoint funds half and gets a $300K operator fee back — net $1.2M 15:34. Smart Lake 27% (Cameco 73%); Hook Lake 21% (Orano and Cameco split the rest) 15:55.
16:44 Partners as validation — and a size filter
- Majors only fund targets that could be tier-one: 100–150M lb. At Hook Lake, once the Spitfire deposit looked like 10–20M lb "tops," they moved on 17:05 — "they keep us honest."
- "In very few juniors do you have to stand up and justify your programs to a major" 17:29 — validation for investors who can't judge the geology themselves.
18:33 Jurisdiction and ore grade
- Saskatchewan has ranked at or near the top mining jurisdiction for a decade (mature permitting, regulation, First Nations relationships). Canada supplies 20–25% of world uranium; Kazakhstan ~40% 19:02.
- Uranium is typically mined at ~0.5%; the two biggest Basin mines run ~20%, Hurricane 45% 19:58. Mining at 2% vs 20% means 10× the rock and 10× the tailings 20:39; a 0.5% find can't just be trucked to McClean Lake — "it's always going to be high-grade or nothing up there" 21:18.
22:22 How the hunt works: structure, alteration, graphite
- Consolidating ~100,000 ha beside Hurricane let them "stand back and get a better look at the geology." Uranium moved with heat and oxygen and settled in structures, altering rock as it went — so they look for soft graphitic zones, rock contacts and alteration halos, found with geophysics 23:36.
- Now "90% of what we're doing is drilling" 25:00.
25:23 The Nova zone: chasing a structure
- Four summer holes hit up to 8%; a bog blocked follow-up until winter, when nine more holes stepped out ~1 km and worked back, finding mineralization along the structure 26:04. A helicopter-supported program is back on it now 26:29.
- Most Basin holes find nothing; >0.05% gets followed up, ≥0.5% is "a significant discovery" 27:49. A typical deposit is Statue-of-Liberty sized, buried half a mile down.
30:00 Not darts at a map — but geology can't be controlled
- Geophysics can eliminate ~90% of a property; targets are ranked and re-prioritized after every hole.
- "If there is no uranium on my project, then I'll never find it" 31:50 — so improve the odds: many targets, digest the data between holes, and rely on people who have seen the rock before. Investors should know "when the jig is up" and when releases turn "more promotional than actually informational" 33:00.
33:35 The press-release study: the market is more efficient than the juniors
- They pointed AI at five years of press releases from 40 Basin companies and their daily prices. Promotional language — "off scale" readings, "composite mineralization" — is common 33:57.
- Findings 35:03: a lot of volume on exciting news is holders exiting; 60% of releases quoting counts per second never followed up with assays 35:49; across ~650 releases, whatever the adjectives, the day-one noise is "all lost" by day three and the stock is "where it's supposed to be" by day five 36:44. "When the news comes out and it excites you… give it a day."
3. In plain English
PUR.V — Purepoint Uranium Group Positive
Purepoint is a small uranium exploration company working in the Athabasca Basin of northern Saskatchewan, and Frostad is its CEO — so this is the company describing itself. Explorers don't produce anything yet; they spend money drilling in the hope of finding a deposit big enough to become a mine, which is why they can multiply in value or go to zero.
Purepoint's twist is that it rarely explores alone. Its main projects are joint ventures with much bigger companies — Cameco, Orano and IsoEnergy — and Purepoint runs the work. That means partners pay a big share of the drilling bills, so Purepoint has to sell fewer new shares (less "dilution" of existing owners), and it earns a management fee for running the programs: on a $3 million season it pays $1.5 million and gets $300,000 back. The partners also act as a check — they only fund targets that could become very large mines (100–150 million pounds), and when one prospect looked too small they walked away. The newest hope is the "Nova zone" next to IsoEnergy's ultra-high-grade Hurricane deposit, where early holes hit up to 8% uranium.
Uranium — the commodity Positive
Uranium is the fuel for nuclear reactors. Demand is easy to forecast: once a reactor is switched on it needs a steady, known amount of fuel for decades. The problem is supply. For several years the world has mined 20–30% less uranium than reactors burned, covering the gap with stockpiles built up after Japan shut its reactors following Fukushima. Those stockpiles are running down.
Normally a higher price quickly brings more supply. Not here: finding a deposit takes years, and turning it into a working mine takes years more — about a decade in total. Meanwhile demand is rising, as tech companies sign power deals for their data centers and new small reactors arrive. Frostad's conclusion is that this "structural supply gap" can't be fixed quickly, so prices should rise sharply.
SRUUF — Sprott Physical Uranium Trust Neutral
You can't legally keep uranium at home, so the simplest way to own the metal itself is through the Sprott Physical Uranium Trust: investors buy units, and the trust uses the money to buy real uranium and store it. Its value tracks the uranium price with no mining risk — the top, safest rung of Frostad's ladder, with the least upside compared with the companies further down.
NXE — NextGen Energy Positive
NextGen is a "developer": it has found a large Athabasca Basin deposit and is working to turn it into a mine. Frostad uses it as the poster child for why explorers are worth the risk — it went from a roughly $40 million company to around $10 billion in a short time once its discovery proved out. Developers like NextGen and Denison are the companies the market is counting on to close the supply gap.
ISOU — IsoEnergy Positive
IsoEnergy made the Basin's most recent discovery, Hurricane: just under 50 million pounds at an average grade of 45% — meaning nearly half of each tonne of ore is uranium, when most of the world mines at about half of one percent. Higher grade means far less rock to dig and process for the same uranium. IsoEnergy is also Purepoint's 50/50 partner in a big land package right beside Hurricane.
CCJ — Cameco Positive
Cameco is the biggest Western uranium miner, which Frostad places on the "producer" rung — the listed companies whose profits respond most directly to a higher uranium price, since they already have uranium to sell. It is also Purepoint's partner on two projects (Smart Lake and Hook Lake), paying most of the exploration bills there.
Editorial summary of the public GG Podcast (Gary Gill) interview with Purepoint Uranium Group CEO Chris Frostad published 16 July 2026 (video linked above); the guest talks his own company. For personal study — not investment advice.