nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqggv8a65afqpcvq2yly8ayjg7j47sl76xcnz86q0pxn9x7m3tfmgsdsds5l (nprofile…ds5l) nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpq8yve5f33hyqq35t3y3xu72t9x33pfdznd0kq78lf73wkxchzr5hs8nh9qy (nprofile…h9qy) I'm not convinced, do you have anything I could read about this (wrt humans)?
When I learned AI at uni, it was about producing new true/false statements from an existing knowledge base using algebraic rules (modus ponens and its friends), so at first I didn’t think LLMs could be considered AI. Since then I reevaluated, considering:
1. that the majority of my knowledge doesn't come from applying logic or from "first principles" at all, but from communicating with other people
1.a there's even an additional parallel here since I apply "weights" to information depending on how much I trust the source
2. that I learned many reasoning tasks ("proof by induction") by seeing examples transmitted to me almost exclusively through language
3. that teaching often involves analogies and figures of speech, which are a feature of language, and often work better than raw logic
To be clear I'm not trying to convince you, I'm just explaining what I thought so you can point me in the right direction!
economística /
npub1fc…wp0wn
2025-04-17 22:28:43
in reply to nevent1q…txxe
Author Public Key
npub1fc3de4qruwq0np7qupcw4w3qytfx3h55j0zdfsmlekt4ntz76qespwp0wnSeen on
wss://relay.nostr.bandPublished at
2025-04-17 22:28:43Event JSON
{
"id": "db3a51817837a8da0a054a1da09d2ff9bb985b52b102e84813c125f7362ccc1f",
"pubkey": "4e22dcd403e380f987c0e070eaba2022d268de9493c4d4c37fcd9759ac5ed033",
"created_at": 1744928923,
"kind": 1,
"tags": [
[
"p",
"42187eea9d4803860144f90fd2491e957d0ffb46c4c47d01e134ca6f6e2b4ed1",
"wss://relay.mostr.pub"
],
[
"p",
"39199a2631b90008d171244dcf2965346214b4536bec0f1fe9f45d6362e21d2f",
"wss://relay.mostr.pub"
],
[
"p",
"beee28c6450eb5d21bf94279d1f3f41e524c87abf86393835b45e6a1e98390e7",
"wss://relay.mostr.pub"
],
[
"p",
"e9c5be23d771931fb9b1e2978b4f5ec915476159fb4dd8c0b1ce17c3da5faaf9",
"wss://relay.mostr.pub"
],
[
"e",
"c597df374e66d37cb759d955ecfa2006a7cd681be9629d0445008a8bec6db914",
"wss://relay.mostr.pub",
"reply"
],
[
"proxy",
"https://mathstodon.xyz/users/dancer_storm/statuses/114355661934220821",
"activitypub"
],
[
"client",
"Mostr",
"31990:6be38f8c63df7dbf84db7ec4a6e6fbbd8d19dca3b980efad18585c46f04b26f9:mostr",
"wss://relay.mostr.pub"
]
],
"content": "nostr:nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqggv8a65afqpcvq2yly8ayjg7j47sl76xcnz86q0pxn9x7m3tfmgsdsds5l nostr:nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpq8yve5f33hyqq35t3y3xu72t9x33pfdznd0kq78lf73wkxchzr5hs8nh9qy I'm not convinced, do you have anything I could read about this (wrt humans)?\n\nWhen I learned AI at uni, it was about producing new true/false statements from an existing knowledge base using algebraic rules (modus ponens and its friends), so at first I didn’t think LLMs could be considered AI. Since then I reevaluated, considering:\n\n1. that the majority of my knowledge doesn't come from applying logic or from \"first principles\" at all, but from communicating with other people\n1.a there's even an additional parallel here since I apply \"weights\" to information depending on how much I trust the source\n2. that I learned many reasoning tasks (\"proof by induction\") by seeing examples transmitted to me almost exclusively through language\n3. that teaching often involves analogies and figures of speech, which are a feature of language, and often work better than raw logic\n\nTo be clear I'm not trying to convince you, I'm just explaining what I thought so you can point me in the right direction!",
"sig": "41311d14514d64d5bc28be760b56db7d2ca874423a0db0efe621b46e78519969f13c331be1b8fb2b6baa72c4f8d4f9d64ff648dfa69646685e5f1f159a788dfe"
}