SpaceX Borrowed $40 Billion for AI Chips and Toyota Is Spending $6.4 Billion a Year on Robots. Who's Winning the AI Race?

On October 25, 2026, Elon Musk said SpaceX had closed $40 billion in debt financing to buy GPUs, the chips that train and run AI models. Reuters had earlier reported that SpaceX was in talks with banks and asset managers to raise the money for Nvidia AI chips. Musk described it as mainly callable debt for GPU purchases. The chips are expected to arrive in the first half of 2027, so the deal won't move this quarter's numbers.

$40 billion: SpaceX's debt raise for AI chips, one of the largest single financing deals for compute ever reported.

Because SpaceX makes real money outside AI. Musk credited its investment-grade status and revenue from Starlink and rocket launches. According to Reuters, SpaceX's third-quarter revenue rose 161% year over year to $14.58 billion. CFO Bret Johnsen said the company expects to spend $35 to $40 billion on capital expenditures in the fourth quarter alone.

Neoclouds are companies whose main business is renting out AI computing power. Musk says there are more than 200 of them. He says debt markets have almost completely stopped lending to companies that don't have strong revenue from something other than AI data centers. His prediction: severe failures, consolidation and only a few survivors. That matters for startups and developers who rent GPUs from smaller providers. If your AI product depends on a single neocloud, a shakeout could mean price jumps or a provider disappearing.

Toyota is making a bet on the physical side of AI. According to Toyota's own estimates reported in September, the group plans to deploy about 400,000 robots, including humanoids, across its factories, group companies and major suppliers. Spending of about $6.4 billion a year is expected to start in 2028. Over a third of the robots are reportedly meant for Toyota Motor Corp. itself. On October 26, the Financial Times reported Toyota's wider ambition to become a global robotics company focused on humanoids and factory automation. Toyota's AI robotics spinout, Walden Robotics, went public in July at a $1.1 billion valuation.

400,000 robots: Toyota's planned deployment across its factories and suppliers.

Toyota CTO Hiroki Nakajima says the robots can learn human movements through cameras and sensors, so skills transfer without complex programming. One example model, ELEY, is a two-armed machine trained by workers wearing an exoskeleton that sends their movements to the robot. This is called teleoperation. Toyota says the robots will augment its workforce, not replace it. The main reasons are worker shortages and passing on skills from experienced staff.

Toyota isn't alone. Automakers are turning into robot makers.

Goldman Sachs analysts think Toyota could take 3 to 8 percent of the humanoid robot market by 2035. Some investors are cautious, pointing to Toyota's slow results with hydrogen cars and solid-state batteries.

So who is actually winning the AI infrastructure race? The winners share one thing: money that doesn't depend on AI. SpaceX has rockets and Starlink. Toyota has the world's biggest car business and its own factories to test robots in. Both can afford to wait years for their bets to pay off. The losers are likely to be companies that only sell AI capacity, borrowed heavily and have no other income. That's exactly the group Musk is warning about.

The big-money AI bets

• Developers: avoid depending on a single small GPU provider; keep a backup
• Startups: expect compute prices to shift as neoclouds consolidate
• Investors: watch for revenue outside AI when judging AI infrastructure companies
• Factory workers: skills in robot training, teleoperation and maintenance are becoming valuable
• Everyone: the next phase of AI is physical, not just chatbots