Running local inference uses more electricity than when at i...

Raw JSON

{
  "kind": 1,
  "id": "3616447838b762ee2ec9481b720f0a3a96f991460451291af02a1215e2644cb6",
  "pubkey": "e83b66a8ed2d37c07d1abea6e1b000a15549c69508fa4c5875556d52b0526c2b",
  "created_at": 1772709551,
  "tags": [
    [
      "e",
      "8480c0b74e1bfb7f2b659ac149f344a3da1d098b80edda652a9a23dfbd371cdb",
      "wss://relay.damus.io/",
      "root",
      "e83b66a8ed2d37c07d1abea6e1b000a15549c69508fa4c5875556d52b0526c2b"
    ],
    [
      "e",
      "016b4cb7994ff0b63dab84001500711b095e18afdc73dcc169425e64b837a262",
      "wss://filter.nostr.wine/",
      "reply",
      "5e5fc1434c928bcdcba6f801859d5238341093291980fd36e33b7416393d5a2c"
    ],
    [
      "p",
      "5e5fc1434c928bcdcba6f801859d5238341093291980fd36e33b7416393d5a2c"
    ]
  ],
  "content": "Running local inference uses more electricity than when at idle, but just like an all you can eat buffet lures you in to a false sense of security, so does running local models.\n\nMost of the time, the cost of local inference hardware costs more than paying for cloud, although API costs can skyrocket.\n\nLocal inference is under your control, however, and the value of that is priceless.",
  "sig": "f0d97b8f80427c7efcbc21d2e7d1ad7173089690c9e15d41e380a5624e4645db5386968fbf3490e128eb0d09dc02124f8bc76b8e3f6a3cf11271db4fb3822d83"
}