Elon Musk – "In 36 months, the cheapest place to put AI will be space”
Source
youtube.com
Author
Dwarkesh Patel
Date
Why it matters
Lays out the power-supply argument for where future AI compute may be built, relevant to long-term inferenceRunning a trained model to get answers — the phase where AI is actually used, as opposed to trained.Full definition → cost and availability, with pushback on GPU servicing and depreciation.
Terms in this piece · Glossary
inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
