2025-10-24 17:32 UTC
@nev @anildash yes, all of this.
There was a thing called "web literacy" that was very popular a few years ago. Crash Course did some episodes on it, but it seems to have gone by the board with too many people apparently being of the opinion that using chatgpt and a web browser are the same thing (aided and abetted by google, it seems).
The point I got from Crash Course on web literacy was that you *use the hyperlinks* to assess the reliability of your source's *site* by seeing what else they promote (it being easier to do this than to assess the accuracy of the webpage you're reading, which would entail finding other webpages that talk about the same thing).
Replies (1)
-
@nev@status.nevillepark.ca 2025-10-24 17:47
@nxskok@cupoftea.social @anildash@me.dm yes! like—I think this gets to the heart of how people come to think LLMs are conscious—just because there are pixels on your monitor forming the shape of letters saying, idk, "such-and-such a plant is toxic to cats," and you read and understand the words, does not mean the plant is actually toxic to cats. like, is it on the website of a university's veterinary medicine school? a page by a botanist specializing in this plant? a Buzzfeed listicle? an SEO-infested, machine-generated page? is the information 20 years old or 2 years old? this actually makes a difference!!!