SpaceX is spending enormous amounts of money right now — on rockets, AI, satellites, and software — and losing billions in the process. At the same time, Wall Street analysts are predicting it will be worth trillions by 2030. Both things can be true. The reason comes down to one simple idea: sometimes you have to spend big before you can earn big.
The "Pay Now, Profit Later" Idea
Think about opening a restaurant. Before you serve a single meal, you spend months paying for the building, the kitchen equipment, the staff training, and the marketing. Your bank account goes down before it ever goes up. That dip — before the profits kick in — is what economists call a J-curve. It looks like the letter J: things get worse before they get better.
SpaceX is right in the middle of that downward dip. The company made $18.7 billion in revenue in 2025 — but still lost $4.9 billion.[2] That's not because the business is failing. It's because SpaceX is pouring money into things that won't pay off for years.
Imagine you're building a toll road. You spend years and billions laying the tarmac, building bridges, and installing systems. You're losing money the whole time. Then one day you open the gates — and the money flows in for decades. SpaceX is still laying the tarmac.
What SpaceX Is Actually Spending On
Most people know SpaceX builds rockets. But the company has quietly become something much bigger — and much more expensive. Here's where the money is going right now:
- Starship, the world's biggest rocket: SpaceX has spent over $15 billion developing it — including $3 billion in 2025 alone.[5] It's designed to be fully reusable, like an aeroplane.
- Grok, an AI chatbot: SpaceX absorbed Elon Musk's AI company xAI in February 2026. The AI division spent $12.7 billion in 2025 and lost $6.4 billion — on just $3.2 billion in revenue.[4]
- Starlink, satellite internet: The one part of SpaceX that's actually profitable — $4.4 billion in operating profit in 2025.[2]
- Orbital data centres: SpaceX has signed a deal with AI company Anthropic worth $1.25 billion per month to test running AI computing from space.[6]
The $322 Billion Question
Goldman Sachs predicts that SpaceX's AI business will generate $322 billion in revenue by 2030 — up from just $3.2 billion today. A 100-times increase in five years.[3]
"The $322 billion number isn't wrong. It's just measuring the finish line — not where SpaceX is standing right now."
AI revenue by 2030. Assumes Starship becomes cheap to fly, Grok becomes a major AI platform, and space-based computing takes off.
Fair value estimate. Morningstar says the AI business faces real competition and the space computing idea is unproven.[8]
Why This Is Harder Than Building Rockets
The AI part of SpaceX's business is actually harder to value than the rocket part. Rockets are physical things you can see and count. AI is mostly invisible — it lives in software, data, and the knowledge inside a company's teams.
Physical assets
You can see a launch pad. You can count launches. A rocket lasts 10–20 years. The value is visible.
Invisible assets
AI models go out of date in 2–3 years. The value only shows up when customers pay for it — which hasn't happened yet at scale.
SpaceX is spending billions on things that don't show up as valuable assets on its books — yet. That's why the losses look so big, and why the future profits look so far away.
Three Signs the Tide Is Turning
- 1Spending slows down relative to earningsRight now SpaceX spends about $40 billion a year but only earns $6.6 billion in operating profit.[2] When those numbers get closer together, the investments are starting to pay off.
- 2Old investments start earning moneyWhen the money coming in from past investments exceeds the money going out on new ones, the corner has been turned.
- 3Revenue grows without hiring more peopleIf Starlink and AI revenue multiplies while headcount stays flat, SpaceX has built something that scales.
What Could Go Wrong
- The space computing idea might not work. Morningstar puts the chance of it failing at 43%.[8]
- Grok faces fierce competition. ChatGPT, Gemini, and Claude are all fighting for the same customers.
- The Anthropic deal is short-term. It's only a 180-day trial with a 90-day cancellation clause.[6]
- Starship still needs to prove itself. Until it's flying reliably and cheaply, the economics don't work.
Standard Chartered has warned that the sheer scale of SpaceX's stock market listing could strain the market.[7] The hype is real. So is the uncertainty.
The Bottom Line
The $322 billion AI forecast isn't fantasy — but it's also not guaranteed. Right now, the company is deep in the "pay now" part of the deal. The "profit later" part is still being built.
The spending years
Billions going out. Losses on the books. Starship still proving itself. AI still finding customers.
The payoff — if it works
Cheap Starship launches. Grok earning real revenue. Goldman projects cash flow swinging to positive $72B by 2031.[3]
We can see the Starships launching. We just can't see the profit yet. That gap — between the spending and the earning — is the whole story.
Sources
- [1]Brynjolfsson, E., Rock, D. and Syverson, C. (2021) 'The Productivity J-Curve', American Economic Journal: Macroeconomics, 13(1). aeaweb.org
- [2]Bradbury, R. (2026) '6 charts: SpaceX's S-1 financials', Yahoo Finance / PitchBook, 20 May. finance.yahoo.com
- [3]Reuters (2026) 'Goldman Sachs expects SpaceX's AI revenue to surge 100-fold by 2030', Yahoo Finance, 4 June. finance.yahoo.com
- [4]Lopopolo, A. (2026) 'SpaceX files IPO paperwork with SEC', Quartz, 20 May. qz.com
- [5]Grush, L. (2026) 'SpaceX Has Spent More Than $15 Billion on Starship', Yahoo Finance / Bloomberg, 20 May. finance.yahoo.com
- [6]Munene, K. (2026) 'SpaceX IPO Nears as Musk Rejects Valuation Cut Report', Analytics Insight, 1 June. analyticsinsight.net
- [7]Khan, W. (2026) 'Standard Chartered says SpaceX, Anthropic, OpenAI IPOs will weigh on stock market', Invezz, 3 June. invezz.com
- [8]Morningstar (2026) 'SpaceX valued at just $780 billion by Morningstar', Yahoo Finance. finance.yahoo.com
This article is for general information only and does not constitute financial or investment advice.
