Sam Altman says Elon Musk's idea of putting data centers in space is 'ridiculous'

SpaceX is pursuing an ambitious plan to deploy a "constellation of a million satellites" as orbital data centers, hiring engineers and pointing to its xAI acquisition to accelerate deployment, while Google’s Project Suncatcher targets space-based, solar-powered data centers possibly as early as 2027. OpenAI CEO Sam Altman called orbital data centers “ridiculous” for the current decade, citing launch costs and repair challenges, even as terrestrial data-center approvals in the U.S. topped roughly 1,200 by end-2024 and spark ESG and local resistance concerns.

Analysis

Market structure: Orbital data centers would create winners among launch providers, satellite manufacturers, high-reliability spacecraft component suppliers and insurers, and losers among terrestrial data-center REITs (EQIX, DLR) and local utility-heavy hosts if adoption gains traction. If launch costs fall toward ~$500–$1,000/kg and on-orbit servicing robotics scale by 2028–2032, pricing power shifts from land-constrained colo providers to vertically integrated space/cloud players (SpaceX/xAI/Google), compressing terrestrial margins by 200–500bps over several years.

Risk assessment: Tail risks include catastrophic orbital debris events, cross-border regulation (ITU/FCC), and insurance losses that could halt deployments — low probability but >$10bn systemic impact. Near-term (days–months) impact is negligible; medium-term (6–24 months) depends on demo milestones (Starship reuse cadence, Google pilot 2027); long-term (3–10 years) is structural if technical thresholds (launch cost, repairability, power delivery) are met.

Trade implications: Tactical longs: GOOGL exposure to Project Suncatcher and cloud (consider 2–3% position) and strategic longs in satellite suppliers (MAXR 1–2%, LHX 1%) vs shorts in EQIX/DLR (reduce weights by 2–4%) as a relative-value pair if launch-cost curve accelerates. Use 12–36 month LEAP calls on GOOGL (e.g., Jan 2028 ~10% OTM) to lever optionality around 2027 pilot timing; buy protective puts on DLR/EQIX if holding short exposure.

Contrarian angles: Consensus underrates operational & regulatory friction — physical repairability and latency for certain workloads make full substitution unlikely this decade, so a complete collapse in terrestrial demand is overdone. The mispricing window is medium term: if Starship achieves <5 flights/month and insured payload loss rates <2% by 2027, re-rate satellite suppliers and Google upward; until then prefer selective, capped exposure and relative trades over outright sector bets.

Key Decisions for Investors

• Establish a 2–3% long position in GOOGL (GOOGL) sized to thesis risk; complement with Jan 19, 2028 LEAP calls ~10% OTM to capture optionality around Google's 2027 pilot milestones and potential re-rating if demonstrators succeed.
• Reduce exposure to data-center REITs EQIX and DLR by 2–4% of portfolio weight over next 3–6 months; if either trades up >8% without supporting Starship/launch-cost progress, layer into short positions sized to freed capital.
• Initiate a long/short pair: long MAXR (1–2%) and LHX (1%) vs short EQIX (1–2%) to play supply-chain upside versus terrestrial colo downside; reassess after Starship achieves a consistent <$1,000/kg benchmark or after Google pilot announcement (target 2027).
• Buy 6–12 month protective puts on DLR/EQIX (strike ~5–7% OTM) sized to current REIT exposure to hedge regulatory/community opposition and near-term demand shocks expected over the next 12 months.
• Monitor specific catalysts for position adjustments: Starship monthly launch cadence (threshold >2 successful flights/month), Google Project Suncatcher pilot contract award (expected by 2027), and FCC/ITU filings in next 30–90 days — if two catalysts trigger, increase space/supplier longs by 50%.