LLMs are good for verifiable tasks where you only want to prove or disprove something, without caring about the intermediate.
It’s a less dumb brute force search, with guidance
semisol
npub122…cgrkj
2026-11-03 14:11:29 UTC
in reply to nevent1q…m5lg
Author Public Key
npub12262qa4uhw7u8gdwlgmntqtv7aye8vdcmvszkqwgs0zchel6mz7s6cgrkjPublished at
2026-11-03 14:11:29 UTCEvent JSON
{
"id": "4bb4f3e4189a4b6b625cc6f6ff09eb1b2f5545dad23561309c09d6fb0ff40ff1",
"pubkey": "52b4a076bcbbbdc3a1aefa3735816cf74993b1b8db202b01c883c58be7fad8bd",
"created_at": 1793715089,
"kind": 1,
"tags": [
[
"e",
"a61dcb943242cce56816c97d8b6a8bbbed8a922579b3e18764af2e5bfc69ab80",
"wss://aggr.nostr.land",
"root",
"abaa09dab1fc3fffc1409fdb28b13832cd4031704c4df6198525fa7ea466bed2"
],
[
"p",
"abaa09dab1fc3fffc1409fdb28b13832cd4031704c4df6198525fa7ea466bed2"
],
[
"client",
"Damus"
]
],
"content": "LLMs are good for verifiable tasks where you only want to prove or disprove something, without caring about the intermediate.\n\nIt’s a less dumb brute force search, with guidance",
"sig": "8ed98c03a568b32519cc8c154bf4c191f39989797c43c8b2b5b992b4d1c7dfef5685beec93f986f10ff690927ebce68f14a9d67099257f6ae6ec051fc852c18a"
}