I came across a claim that a query to an LLM is no worse than watching a Youtube video. This is not correct. Serving streaming video is surprisingly resource efficient. Let's have a look. (1/5)
#FrugalComputing
Wim🧮 /
npub1wa…r47x2
2025-01-07 11:54:49
Author Public Key
npub1waphul5nh8y8aqhpagp9vvx8azayfzk4ftk4q5n4a30xckz2mqxsxr47x2Published at
2025-01-07 11:54:49Event JSON
{
"id": "aebae66db1b589396999a4edcacd2e5cd7a4c678023f997f96beae5a06097490",
"pubkey": "77437e7e93b9c87e82e1ea025630c7e8ba448ad54aed505275ec5e6c584ad80d",
"created_at": 1736250889,
"kind": 1,
"tags": [
[
"t",
"frugalcomputing"
],
[
"proxy",
"https://scholar.social/users/wim_v12e/statuses/113786938293256655",
"activitypub"
]
],
"content": "I came across a claim that a query to an LLM is no worse than watching a Youtube video. This is not correct. Serving streaming video is surprisingly resource efficient. Let's have a look. (1/5)\n\n#FrugalComputing",
"sig": "8d7fb4f751f6cd3c722d7e665105d00e1b144ac137fe0e4847e46521cfe295fe690e67bfbc2c5bda6e6530eb961c12c19ecbef7f6eb8962586f25c7c460f50a1"
}