“What the Apple paper shows, most fundamentally, regardless of how you define AGI, is that LLMs are no substitute for good well-specified conventional algorithms.”
https://garymarcus.substack.com/p/a-knockout-blow-for-llms
zeldman
npub1zp…kje3a
2025-06-11 14:06:03 UTC
Author Public Key
npub1zpydpzt5gvxqr0dg2zj4kal6r87z69n2ecmrwm9v7vt576nlt0zqskje3aSeen on
wss://relay.momostr.pinkPublished at
2025-06-11 14:06:03 UTCEvent JSON
{
"id": "44d531123807e76fdc0b7c363cb38328145677efe5d3ce5499e476e48fafba45",
"pubkey": "1048d08974430c01bda850a55b77fa19fc2d166ace36376cacf3174f6a7f5bc4",
"created_at": 1749650763,
"kind": 1,
"tags": [
[
"proxy",
"https://front-end.social/@zeldman/114665112452760728",
"web"
],
[
"proxy",
"https://front-end.social/users/zeldman/statuses/114665112452760728",
"activitypub"
],
[
"L",
"pink.momostr"
],
[
"l",
"pink.momostr.activitypub:https://front-end.social/users/zeldman/statuses/114665112452760728",
"pink.momostr"
],
[
"-"
]
],
"content": "“What the Apple paper shows, most fundamentally, regardless of how you define AGI, is that LLMs are no substitute for good well-specified conventional algorithms.”\n\nhttps://garymarcus.substack.com/p/a-knockout-blow-for-llms",
"sig": "e8e732e267e5ddb46dd6d8177b9b1e5774aef41f0e50d90013a0aadfc1c5d96247d766107cb11575e9e343821bdcb8d34e8eb25bcc5bdf030c2c5b211ff4dde3"
}