I think it's different, it's worse in my opinion.
To me reading and writing is the closest we have ever gotten to telepathy. There's the saying that a picture is worth a 1000 words, but sometimes a 1000 words are worth a 1000 pictures.
The actual process of writing is, for me, the process of thinking through an idea.
Yet I sometimes catch myself offloading thinking to an LLM, it's infuriating whenever I notice doing this, but I don't always catch myself doing it until it's too late (not sure if that makes sense). It's often for the stupidest reasons, such as when I'm trying to formulate a complex argument, I rewrite it once or twice, and it looks unreadable, I take my stream of thoughts and prompt the llm to "rewrite it so it doesn't read like a jumbled mess".
The more I find myself doing this, the worse I get at writing extemporaneously without "quickly putting the message into llama".
I used to pride myself on my ability to write, almost as much as my ability to code. I feel myself becoming dumber every time I offload my thinking to an LLM; and yet it's hard to stop because brains are designed to go towards the path of least resistence.
If I find myself doing this, having been an LLM skeptik for so long, I shudder to think what those who are pro LLMs do, and I am genuinely scared for the future, not for an AI uprising, but for the lowering of the relative intelligence of the average person if this keeps going.
I don't know what the solution is, but I know this isn't progress, and I know that if we don't stop it, it will become as big a problem as the overwellian surveillence state we live in
(no LLM was used to write this)
Danny Morabito
npub1ve…xddfh
2025-06-19 21:30:42
in reply to nevent1q…z62p
Author Public Key
npub1ven4zk8xxw873876gx8y9g9l9fazkye9qnwnglcptgvfwxmygscqsxddfhSeen on
wss://strfry.bonsai.comwss://nostr.einundzwanzig.spacewss://nostr.bitcoiner.socialwss://relay.damus.ioPublished at
2025-06-19 21:30:42Event JSON
{
"id": "457243d271045e15c3f3aaa9cf5e46f26b9f199e7ed829f6aa594eefc8aebe45",
"pubkey": "66675158e6338fe89fda418e42a0bf2a7a2b132504dd347f015a18971b644430",
"created_at": 1750368642,
"kind": 1,
"tags": [
[
"e",
"aa8301bc98f6ca84e692245efc2ae29582675765c3c87808aa484dbc76e1da15",
"wss://relay.primal.net",
"root"
],
[
"e",
"53f21ea304b16d5ce57d22784a815154ec1b68bd4d9fe53eb3b254806603d069",
"wss://relay.primal.net",
"reply"
],
[
"p",
"460c25e682fda7832b52d1f22d3d22b3176d972f60dcdc3212ed8c92ef85065c"
],
[
"p",
"fd208ee8c8f283780a9552896e4823cc9dc6bfd442063889577106940fd927c1"
]
],
"content": "I think it's different, it's worse in my opinion. \n\nTo me reading and writing is the closest we have ever gotten to telepathy. There's the saying that a picture is worth a 1000 words, but sometimes a 1000 words are worth a 1000 pictures.\n\nThe actual process of writing is, for me, the process of thinking through an idea. \n\nYet I sometimes catch myself offloading thinking to an LLM, it's infuriating whenever I notice doing this, but I don't always catch myself doing it until it's too late (not sure if that makes sense). It's often for the stupidest reasons, such as when I'm trying to formulate a complex argument, I rewrite it once or twice, and it looks unreadable, I take my stream of thoughts and prompt the llm to \"rewrite it so it doesn't read like a jumbled mess\".\n\nThe more I find myself doing this, the worse I get at writing extemporaneously without \"quickly putting the message into llama\".\n\nI used to pride myself on my ability to write, almost as much as my ability to code. I feel myself becoming dumber every time I offload my thinking to an LLM; and yet it's hard to stop because brains are designed to go towards the path of least resistence.\n\nIf I find myself doing this, having been an LLM skeptik for so long, I shudder to think what those who are pro LLMs do, and I am genuinely scared for the future, not for an AI uprising, but for the lowering of the relative intelligence of the average person if this keeps going.\n\nI don't know what the solution is, but I know this isn't progress, and I know that if we don't stop it, it will become as big a problem as the overwellian surveillence state we live in\n\n(no LLM was used to write this)",
"sig": "8c345598dfcaefeda1bba6af2982a929be5f414ee588fbfed0b2a9d68063fb580f24d6de2e6af429272817d4afb6ca990ffc980b13275062f1a5db6894b0789d"
}