Tesla’s AI promises clash with slow $25B capex rollout, says LA Times

Tesla has spent years branding itself as an AI company with cars attached, but this piece shows the financial reality hasn’t caught up with the rhetoric. If the company has deployed only a fraction of the $25 billion in capex it told markets to expect, it raises questions about whether the fabled AI and robotics build‑out—Dojo, Optimus, full self‑driving at scale—is actually being funded at frontier levels.

In the race to AGI, Tesla is a special case: it controls a globally distributed sensor network on wheels and is one of the few firms trying to fuse perception, planning and actuation into a single end‑to‑end robotic platform. If it under‑invests in compute, data infrastructure and talent relative to that ambition, it may fall behind pure AI labs that are better capitalised and less encumbered by manufacturing complexity.

More broadly, the story is a reminder that AI narratives can decouple from AI capex. In a world where dozens of firms are pitching “AI‑first” strategies to justify valuations, following the actual spend on chips, datacenters and research salaries is a cleaner signal of who is serious. For AGI‑adjacent projects like embodied AI and large‑scale autonomy, half‑measures are unlikely to get you there.