SpaceX to Invest $2B into Elon Musk's XAI(wsj.com)
wsj.com
SpaceX to Invest $2B into Elon Musk's XAI
https://www.wsj.com/tech/spacex-to-invest-2-billion-into-elon-musks-xai-413934de
18 comments
The SpaceX investment is especially problematic as SpaceX has gotten $22bn government contract dollars. So depending on what money he is moving, taxpayers may be footing the bill for Grok development.
They have also saved the govt over $100B compared to paying Russia and Boeing etc.
Are you aware that many companies receive billions in government contracts, and why are you concerned about this and not others e.g. amazon, MSFT, etc
Personally, I don't care because AWS and Azure are both extremely profitable without federal contracts.
XAI is the dictionary definition of a blue-chip nothingburger. Nobody wants their tax dollars going towards the bailout of an AI company that can't compete on merits alone. It's a dilution of SpaceX's valuation, if nothing else.
XAI is the dictionary definition of a blue-chip nothingburger. Nobody wants their tax dollars going towards the bailout of an AI company that can't compete on merits alone. It's a dilution of SpaceX's valuation, if nothing else.
Ah, so your premise is SX is unprofitable without federal contracts! and that the SOTA model, which was released before this investment, can’t “compete”
I think the real underlying premise is that Elon is unhinged, cannot be trusted and just makes stuff up. Self driving Teslas were going to launch "next year" since 2017. DOGE was going to find 1-2T dollars worth of waste in our budgets. Thai rescue diver was a pedophile. etc.
For all I know, SOTA model can be a copy with some additive work on Claude or OpenAI models.
For all I know, SOTA model can be a copy with some additive work on Claude or OpenAI models.
You are exactly right that the underlying premise is a dislike of Musk, who has famously made mistakes, I too am critical of all of the things you listed.
So, if a very contentious personality is involved, is there any purpose or value in pointing out seriously flawed assumptions/POV/rationale?
Another example is the excessive divisiveness in politics, what would make discussions around those topics better? In my opinion, it would be better to rein in serious factual errors, even if the errors skew towards "my side"
So, if a very contentious personality is involved, is there any purpose or value in pointing out seriously flawed assumptions/POV/rationale?
Another example is the excessive divisiveness in politics, what would make discussions around those topics better? In my opinion, it would be better to rein in serious factual errors, even if the errors skew towards "my side"
It's about the ideas, and the idea is bad. It would be bad if Elon Musk came up with it, it would be bad if Martin Luther wrote it down on a handkerchief on the eve of his death.
Do we want to discuss this on merits, or are we concerned that the merit of the idea might undermine... checks clipboard ...commercial LLM businesses cashing R&D checks expensed by the US taxpayer?
Do we want to discuss this on merits, or are we concerned that the merit of the idea might undermine... checks clipboard ...commercial LLM businesses cashing R&D checks expensed by the US taxpayer?
“The idea” is that (the only) megaconstellation sat internet provider is investing in and partnering with an AI company. On its face that seems to make sense to me. If it does not make sense, why not? The reasons you have given are irrelevant and also false.
The “taxpayer” (govt) pays SX a very low rate for launches, and gets an excellent product in return. In what way is the “taxpayer” being swindled?
The “taxpayer” (govt) pays SX a very low rate for launches, and gets an excellent product in return. In what way is the “taxpayer” being swindled?
That's not what I said at all. If AWS or Azure went all-in on a risky, overvalued technology, they would get extreme scrutiny from both the public and federal contractors. SpaceX is doing exactly that, and diluting their valuation in the process. You'll note that neither AWS nor Azure have made similarly risky "balls on the table" ventures.
You can argue that the dilution is good for taxpayers in the long-run (fat chance lol), or that SpaceX can survive the immediate aftermath (duh). But you cannot argue that XAI is a deserving company, because it never won a contract against it's competition. The demand is invented, simple as that.
Mind you, XAI couldn't even fund itself before the seed rounds. Both political admins will be looking at their SpaceX contracts with a renewed scrutiny, which Elon had better hope doesn't develop into a campaign for nationalizing SpaceX IP.
...who am I kidding, he was begging to become eminent domain the moment he put POTUS in his crosshairs. Au revoir, rocket man.
You can argue that the dilution is good for taxpayers in the long-run (fat chance lol), or that SpaceX can survive the immediate aftermath (duh). But you cannot argue that XAI is a deserving company, because it never won a contract against it's competition. The demand is invented, simple as that.
Mind you, XAI couldn't even fund itself before the seed rounds. Both political admins will be looking at their SpaceX contracts with a renewed scrutiny, which Elon had better hope doesn't develop into a campaign for nationalizing SpaceX IP.
...who am I kidding, he was begging to become eminent domain the moment he put POTUS in his crosshairs. Au revoir, rocket man.
Is SX running out of money? They will be if they keep doing this, but I was under the impression they were still insanely profitable.
Really hard to know - as the numbers are not public, it's all guestimates.
Musk said in late 2023 that they achieved 'breakeven cash flow', but Bloomberg reported that they were actually loosing money on ground terminals.
It's also a lot of accounting game - Starlink satellites have a really short lifespan and will need to be replaced on a regular basis. SpaceX being a private company, there is no way to know how they account for that.
As for launches - most of their launches are for... themselves (Starlink) and quite a lot of the rest is for government / military - which are partly classified.
Musk said in late 2023 that they achieved 'breakeven cash flow', but Bloomberg reported that they were actually loosing money on ground terminals.
It's also a lot of accounting game - Starlink satellites have a really short lifespan and will need to be replaced on a regular basis. SpaceX being a private company, there is no way to know how they account for that.
As for launches - most of their launches are for... themselves (Starlink) and quite a lot of the rest is for government / military - which are partly classified.
A few years ago, Musk said that if Starship isn't putting next generation Starlink satellites into orbit, SpaceX will go bankrupt.
5G FWA is eating away at the edges of Starlink's TAM, which makes delays even more expensive.
5G FWA is eating away at the edges of Starlink's TAM, which makes delays even more expensive.
It will be as impressive as it is shortsighted if we throw away a strong lead in space (at least the US chunk of it) due at least partially to political feuding and populism.
I do wish the fellow had stuck with the boring (hah) business of space colonies and robot cars instead of getting sidetracked and baited with other stuff.
I do wish the fellow had stuck with the boring (hah) business of space colonies and robot cars instead of getting sidetracked and baited with other stuff.
Is there any opportunity to integrate or leverage starlink with AI?
Thus undoing hundreds of billions of dollars vis-a-vis risk management in space.
PS. For those about to write that Tesla has 30B in cash - it would be a very nice runaway for a small company, but when you are a carmaker employing 100k people and producing cars that are not selling, that cash pile burns really quickly.