supplierScore 75/100Watch
Hotz cites Google Cloud ToS prohibition on autonomous vehicle training; names Nvidia as monopoly supplier with exploitative pricing.
“Your options are Nvidia or Google. Um, so Google is not even an option. uh their TPUs are only available in Google Cloud. Uh Google has absolutely ownorous terms of service restrictions. Uh they may have changed it, but back in Google's terms of service, it said explicitly you are not allowed to use Google Cloud ML for training autonomous vehicles or for doing anything that competes with Google without Google's prior written permission... that leaves you with the monopoly uh Nvidia.”
Why it matters
This exposes a critical supply-chain chokepoint: autonomous vehicle companies cannot use Google TPUs (restricted by ToS), leaving Nvidia as the only viable training platform. Hotz explicitly calls out Nvidia's monopoly pricing ($10k for A100 vs $699 for RTX 3080) as unsustainable and claims it is pushing customers to build alternatives.
Investment implication
Creates immediate demand signal for alternative AI accelerators and custom silicon. Companies building competing inference/training chips (AMD, Cerebras, Graphcore, custom designs by Tesla/others) now have clearer TAM. Also suggests Nvidia may face customer defection if pricing remains aggressive; however, near-term, Nvidia's monopoly in this segment remains intact.