hagen :birdsite: on Nostr: with nvidia entering the llm space while also being the sole manufacturer with gpus ...
with nvidia entering the llm space while also being the sole manufacturer with gpus good enough to train the high end models, why would anyone still invest billions into openai? even if you believe in a future where these things are actually useful and profitable, the competition is still the only company that doesn’t have to pay the insane markups on the most important part of the whole operation. the math isn’t mathing.
Published at
2024-10-13 12:05:36Event JSON
{
"id": "076d7fc5543063d5daf056977425505df79c6e910d8cd161e9312766c244b2f8",
"pubkey": "bbfb9a204506863337392bd1afc8a8e40ca76588f6f1034926415a0f95e4bb31",
"created_at": 1728821136,
"kind": 1,
"tags": [
[
"proxy",
"https://mastodon.social/users/hagen/statuses/113300022029110653",
"activitypub"
]
],
"content": "with nvidia entering the llm space while also being the sole manufacturer with gpus good enough to train the high end models, why would anyone still invest billions into openai? even if you believe in a future where these things are actually useful and profitable, the competition is still the only company that doesn’t have to pay the insane markups on the most important part of the whole operation. the math isn’t mathing.",
"sig": "3e4d779c531014ca756c84edce6dec7aafb876bf8ec6bf9d4b624e6c10f557572b3eacac991f4713b7048579cfe6c803364ce29b25e6fd31a01a84ab34f37f5c"
}
