Behind the video
The SpaceX Thesis Just KEEPS Getting Better
Overall Thesis
The host argues that expanding AI capabilities, falling costs per task, and growing adoption will sustain enormous compute demand for decades, benefiting SpaceX's compute infrastructure and AI products. He contrasts SpaceX's flexible compute contracts with CoreWeave's fixed pricing, but provides no qualifying price or investment-return prediction for a public asset.
Narratives
CRWVCoreWeaveThe host endorses the criticism that CoreWeave's long-term fixed-price compute contracts sacrifice potential revenue as market rental rates rise. This is a relative business-model disadvantage against SpaceX's contractual flexibility, rather than a prediction that CoreWeave shares will decline.
Key Arguments
- Fixed-price contracts limit participation in higher compute rental rates.
- The host says greater contractual flexibility could have allowed CoreWeave to earn substantially more.
- SpaceX's described 90-day termination provisions provide flexibility to reclaim capacity or potentially renegotiate pricing.
Risks acknowledged
- The discussion credits CoreWeave with operating reliable GPU clusters and bringing compute capacity online effectively.
- The criticism concerns contract economics rather than an inability to supply working compute.
- The broader thesis anticipates substantial growth in AI compute demand.
GOOGLAlphabetThe host cites Google as a major customer willing to pay for access to SpaceX compute, using that relationship as evidence of strong infrastructure demand. He does not establish a directional investment view on Alphabet shares.
Key Arguments
- Google is described as paying roughly a billion dollars per month for compute access.
- The host uses demand from Google and other AI developers to support his thesis that compute supply remains constrained.
Hedges & Caveats
- The host says he is not an expert and describes his compute spot-price outlook as a hunch.
- The proposed repricing of SpaceX contracts is explicitly hypothetical.
- He acknowledges that his long-term SpaceX valuation model will be wrong and does not know in which direction.
- He urges viewers to evaluate the assumptions independently and acknowledges they may reach different conclusions.
- SpaceX is excluded from ticker fields under the supplied private-company restriction.