Zhenyi Tan on Nostr: "Training an LLM with more data won't make it smarter. [...] It's like a search ...
"Training an LLM with more data won't make it smarter. [...] It's like a search engine that has indexed only 100 web pages isn't very useful, but a search engine that has indexed 10 billion web pages will have a much higher chance of returning a relevant result."
https://zhenyi.gibber.blog/ai-advertisingPublished at
2024-09-03 20:50:11Event JSON
{
"id": "b1c75b2dcc3611a21833cc85cc4894e5ab097049fad4337472e2521682eaca2a",
"pubkey": "ba9d498f21902a9836dea1ae3d1dd894e5507155cf229c47a8275ff8152f9a14",
"created_at": 1725396611,
"kind": 1,
"tags": [
[
"proxy",
"https://mastodon.social/users/zhenyi/statuses/113075592360722111",
"activitypub"
]
],
"content": "\"Training an LLM with more data won't make it smarter. [...] It's like a search engine that has indexed only 100 web pages isn't very useful, but a search engine that has indexed 10 billion web pages will have a much higher chance of returning a relevant result.\"\n\nhttps://zhenyi.gibber.blog/ai-advertising",
"sig": "8c0b6cb5ab232268083f3ceb7af81fedbce810da6f3527a0af39055f6759dfb365bdd4893a8e74decdec1c18f769f106ab8380f3b4120f05495a5dd89cc6e975"
}
