Behind the video
History is About to Be Made... It's Worse Than You Think
Overall Thesis
AI is not a bubble despite bear warnings; the strongest financial evidence shows AI companies are delivering exceptional growth, presenting a significant money-making opportunity for investors who understand the real risks.
Narratives
Tom Nash views Constellation Energy as a key beneficiary of the AI power bottleneck, citing nuclear energy as essential for powering data centers going forward. He considers the position 'still early' despite already being up since he added it to his list.
Key Arguments
- Nuclear energy is expected to be a massive staple in powering AI infrastructure.
- Data center electricity demand is expected to quadruple by 2034, creating a power bottleneck that companies like Constellation can help solve.
Tom Nash argues that cooling technology will be critical to solving AI's power/infrastructure bottleneck and names Vertiv as the top pick in this space. The stock is up 111% since he added it to his list in June of last year.
Key Arguments
- Cooling infrastructure is described as critical to solving AI's data center energy bottleneck.
- Vertiv is positioned as a leading solution provider in this space.
BEBloom EnergyTom Nash describes Bloom Energy as a volatile but promising play on on-site power generation for data centers that don't need grid connection. He notes it can swing significantly in either direction on any given day, reflecting high uncertainty despite overall bullish framing of the energy bottleneck thesis.
Key Arguments
- Provides on-site power generation independent of grid connection, addressing the AI power bottleneck.
- Highly volatile, capable of moving 20-30% in a single day in either direction.
Risks acknowledged
- Described as 'very bipolar, very volatile' with potential for large single-day drops.
NVDANvidiaTom Nash argues Nvidia's infrastructure layer (CUDA, hardware) has a durable moat unlike commoditized LLM models, evidenced by 75% hardware gross margins and massive developer lock-in. He believes Nvidia is undervalued given its growth rate and PEG ratio, making it cheap despite being one of the best companies in the world.
Key Arguments
- CUDA has 20 years of development, six million developers, and thousands of optimized libraries creating switching costs.
- 75% gross margin in hardware is only possible with a strong moat and pricing power.
- PEG ratio of 0.5 with a mid-20s P/E makes Nvidia cheap even if growth slows to 50%.
- Trailing 12-month revenue of $250 billion and $160 billion net income show fundamental strength, not a bubble.
Risks acknowledged
- LLM model layer itself is commoditized and has no moat.
- Power/grid bottlenecks could constrain future data center buildout needed for AI compute.
Predictions (1)
"So even if Nvidia maintains 50% growth for the next couple years, which would be below its current growth rate, it's still going to be cheap based on this valuation."
PLTRPalantirTom Nash believes Palantir is priced for perfection but may deserve it given its exceptional financial profile, including 145 Rule of 40, 150% net retention, zero debt, and $8 billion cash. He sees its government contracts and institutional integration as a nearly unbreakable moat.
Key Arguments
- Rule of 40 of 145, 85% revenue growth, 60% operating margin, and 150% net retention are described as a 'one of one' financial profile.
- Government contracts, security clearances, and years of institutional integration make Palantir nearly impossible to replicate.
- Zero debt and $8 billion in cash provide financial stability.
Risks acknowledged
- Trades at a 14x sales-type valuation ('14p') that is priced for perfection with no room for execution missteps.
Predictions (1)
"Piloter is priced to perfection, but they might be as close to it as it gets."
CRWDCrowdStrikeTom Nash briefly cites CrowdStrike's record ARR growth as evidence that leading AI-adjacent infrastructure companies are reporting their strongest quarters despite bearish sentiment in the market. No detailed thesis or prediction is given beyond this data point.
Key Arguments
- CrowdStrike set a new ARR record of 32% growth.
Hedges & Caveats
- Nothing in this video constitutes tax, legal, financial and/or investment advice
- Acknowledges legitimate third argument against AI that bears are not discussing
- Warns that most investors will either panic sell or hold blindly without understanding key risks
- States opinions are subjective ('opinions are like buttholes, everybody's got one')
- Emphasizes need to separate legitimate risks from panic