this guy is asking the wrong question; the right question is not how much money SpaceX is being paid, but why they’re making the deal in the first place: because they’ve realized that they aren’t going to win in the frontier model race (and need things to look good for the IPO).

Whether SpaceX/ Xai is making money on the deals with Google and Anthropic or losing money, they are waving the towel on winning the frontier model race— by arming their competitors rather than themselves.

@GaryMarcus spacex + google collab amazing fr

@GaryMarcus not wrong tbh. lending your edge out usually means u dont trust ur own speed anymore

@GaryMarcus Yes, and? You’re mad because they’re pivoting to something that makes more sense?

@GaryMarcus So what (if even true)

SpaceX is about so much more than this drop of water in the ocean

@GaryMarcus Doing a side thing with 10% of their capacity doesn't count as a pivot

@GaryMarcus Yet, the AI TAM is the largest in the S-1. They always want it both ways.

@GaryMarcus strong take honestly

if youre arming ur competitors, must be a different endgame nobody sees yet

@GaryMarcus Competition breeds innovation. SpaceXAI is training 7 concurrent frontier models in conjunction with these deals with the edge in compute they built over others. They are just helping fund the mission. There isn’t 1 winner in AI.

@GaryMarcus the IPO optics angle is actually the most honest read here

I HATE to break it to you but SPACEx is alot more than just xai lmao🤣 just cause they don't have the best LLM doesn't mean shit when they also have over 600 successful rocket launches and a satellite network capable of providing Internet to damn near the whole world. This isn't only an ai race.

@GaryMarcus I don’t image this deal much outlasts Google’s 12 mo lockup on their $100B stake in spacex. They surely wants to sell SpaceX equity rather than their own (after just doing $80b) to fund their capex

@cunha_tristan fair point that the contracts aren’t long term, but it still doesn’t speak well for their confidence in their original scaling hypothesis if they can’t find something useful to do with 330,000 H100s.

@Robotbeat @GaryMarcus It only makes sense in the short term for IPO purposes

The value agreed is absurd & unsustainable. Google is an investor and benefits far more from maintaining a high share price. This is merely an insurance payment. Google will sell out and cancel the moment its lock-in expires

Key question in any business is who captures the value. It is not that SpaceX is out of the race. They missed the code agent wave (and have a backup plan with Cursor). Given current scenario, there is more money in providing power for the compute. Compute without power is idle GPUs.

@GaryMarcus The race is not a sprint. Elon never gives up. Ever. If you are counting him, you are not very smart.

@GaryMarcus It's a massive strategic trade-off. They're essentially trading compute access for immediate liquidity, but it creates a massive moat for the very companies they're trying to outpace.

@GaryMarcus Gary doesn’t even know Colossus 2 is a reality lol… he thinks BYD had a good Q1 as well 🤦🏻‍♂️. If you work at MIT , you probably would assume he is be good digging into the details, right??

@GaryMarcus Bad take. It’s pretty rationale no? They stood up infrastructure for SOTA LLM and huge inference demand that hasn’t materialized.

They can rent out infra, while training next version of Cursor / xAI, and ultimately reclaim pretty quickly if model is able to get back into T1