If I may bring a bit to the debate, I had this thought.
Authorizing more shares than outstanding is indeed pretty common, as noted by
@Yochana42437421.
But the question is: to do what?
1- Raise capital
2- Buy something (and this something is clearly Tesla... not OpenAI lol)
In reality it will be both, the question is what will be the ratio. We can quantify this if we look at what SpaceX wants to do.
SpaceX will need indeed to raise capital. Once Starship is reusable, mass production will start. And the near term target will be to deploy massive AI compute in Space. Let's first aim at 10x the current terrestrial US compute and let's do the math for how much it would cost.
- SpaceX/xAI plans target ~100 kW of compute power per ton of satellite mass at scale.
- 10x current US compute would mean scaling to ~8โ9 million GPU-equivalents or equivalent FLOPS/power capacity (roughly 3 GW of AI-focused power draw, depending on efficiency).
- At 100 tons per Starship launch: ~300 launches total for the full constellation
- Launches per year per Starship: ~180 (365 days / 2).
- For 300 total launches in one year: ~2 Starships (theoretically, with perfect cadence and rapid turnaround). More realistically: 10 Starships
- Cost of one Starship unit booster today: $90 million. This is at prototype level, if mass production starts, $50 million should be reachable, $500 million to build the fleet. This is peanuts.
- AI tonnage: 100 kW of compute power per ton, 3 GW target, total mass needed: ~30,000 tons
- If we assume a cost of $2000/kg for AI compute, we need $60 billion to build 10x the current US compute in the form of AI satellites.
Therefore, the cost of building and launching 10x the current US compute in Space, including building the Starship fleet, is around $60 billion. Triple that (current US compute is still growing, cost may be higher at first), you are still under $200 billion.
So if you look at the math, there is no way that SpaceX would need this amount of capital. The best explanations are:
- To buy Tesla
- Or just in case for the future... It's not because you can issue these shares that you will issue these shares. This is of course possible.
So if you connect the dots:
- Elon has a super majority at SpaceX, SpaceX will therefore vote yes to any terms
- Tesla shareholders need to be convinced and offering very generous terms is the best way to do this
- Even if Elon offers very generous terms to Tesla shareholders, due to his super voting rights, he still maintains control (Jo's point)
- SpaceX will not need to raise amount of capitals in the order of magnitude implied by the newly authorized shares
It is likely that these newly authorized shares are there to buy Tesla at a super premium.
@JOBhakdi @TeslaBoomerMama @CernBasher
Jo, you are CRAZY.
Let's debate.