[bluegatty]

I'm sorry this is a misplaced framing.

==> Those facilities are being leased because Grok is failing.

Space X does not want to lease away it's competitive advantage to a primary competitor.

It'd be like Tesla leasing factory space to Toyota and Ford.

'GPUs, Energy and Data Centres' are a hugely critical resource in the AI race and SpaceX is now leasing it away.

Will it make money? Sure.

But this is 'Strategic Fumbling'.

The cash flow happens to help them leading up to IPO - that's a side show.

[d0gsg0w00f]

I think they're sacrificing short term terrestrial GPU advantage to fund their space GPU mega-advantage. It's a risky bet, but if it pays off then SpaceX will be one of the most powerful companies the earth has ever seen.

[bluegatty]

It's not that complicated, and the 'datacentres in space' is a myth.

They have rapidly depreciating assets in GPUs and they can't use them.

Because Grok is failing.

They are licensing out their unused capacity.

XAi is not strongly related to Space X - they were folded into one thing because XAi was losing money and failing (the Social Media part is worse).

XAi isn't really some kind of strategic advantage for Space X and even though revenues from the data centres may be positive - it's a 'stop gap' - it's probabaly not a 'net positive' thing to do.

The best thing for Space X would have been to never merge wht XAI.

The second best thing for Space X would have been to close XAi/Twitter lines of business a long time ago.

The 'Wrench in the Logic' is that by putting these things together, EM is able to dupe so many people into so many ridiculous concepts.

Data centres in space, 1M people on Mars, all sorts of crazy things.

It's a bit like Putin's and Steve Bannon's Media Strategy: 'Flood the Zone' with nonsense, and people speculate as to all kinds of things.

The Space X IPO is a 'retail push' meaning he needs to get all the Dentists in America and their Private Bankers to want to 'Hold the Bag' and then keep holding it for a long time.

Space X - at it's core - is a decent company, wrapped in layers and layers of hyperbolic nonsense.

All it takes is a bit of rational thinking to wade through what is plausible, and what is not, and we can see how overvalued this is.

Note: this is different than the other 2 AI IPOs which have some sketchy economics - but the premise is not far fetched, 'that people will want AI in large quantities'.

[poisonfountain]

It is probably both.