Christopher Mims on Nostr: I don't mean to be glib about what LLM-powered chatbots and AI-powered answer engines ...
I don't mean to be glib about what LLM-powered chatbots and AI-powered answer engines could do to the internet.
Research by Ethan Mollick and others shows that when people ask an AI for answers, they *do not click on links* those answers are referenced from.
This is extremely bad for a lot of publishers.
2/🧵
Published at
2024-02-17 13:26:12Event JSON
{
"id": "4dea0c23ddaedd3c6f14106ecf2b4cc3224847446a27106f51ae1b35e45a6ff2",
"pubkey": "e61ff23992c86d2a887a687f5336f76964d5e1ae9b5b3631d8934c431432fb0a",
"created_at": 1708176372,
"kind": 1,
"tags": [
[
"e",
"07d30c1735987fd45679d7c57cbf66981dca92850b8060c33c14376e2b5bf489",
"wss://relay.mostr.pub",
"reply"
],
[
"proxy",
"https://mastodon.social/users/mimsical/statuses/111947046777369197",
"activitypub"
]
],
"content": "I don't mean to be glib about what LLM-powered chatbots and AI-powered answer engines could do to the internet.\n\nResearch by Ethan Mollick and others shows that when people ask an AI for answers, they *do not click on links* those answers are referenced from.\n\nThis is extremely bad for a lot of publishers.\n\n2/🧵",
"sig": "162186cf44953a88609433ff25db9f5bf7d3b74a0120136f4315b75135c8c6929b881de714f9d3a2cf0413f94cc5429147970e9a77c89cb6b0246257e719e1cd"
}
