71–80 of 100 posts
I always think people calling AI just ‘fancy autocomplete‘ haven’t really tried it yet.
On a seperate note, why do people feel the need to exaggerate their Quora profile/education level/accomplishments? I'm not the only one who's noticed how many Quora profiles are filled with ludicrous, self-aggrandizing descriptions, right?
> On a seperate note, why do people feel the need to exaggerate their Quora profile/education level/accomplishments? As in it is Quora specific or it's always been this way? People haven't changed...
I always think people calling AI just ‘fancy autocomplete‘ haven’t really tried it yet.
It's surprising to see gendered, pejorative language thrown around in an article that has nothing to do with gender.
Earlier quoted context omitted.
I found this is a “marketing” technique, that I first encountered on LinkedIn Groups. A shill asks a specific question, which is then “answered” by a bunch of other shills, recommending some product. I remember answering a technical “question” on the Swift Language group, in LI. I gave a simple, accurate, answer to what I considered to be an earnest question about a basic dilemma. It was something that I knew all abo…
Linkedin also goes the quora way. Nope, it's not marketing, it's something else.
It may actually be a technique, taught in school. There's a lot of underhanded behavior in that field, especially in trying to drive talking points and corporate glossaries.
I think, it's not AI that ruined Quora, but monetization. It's a strong case of Goodhart's Law. The measure is popularity and user retention.
In an ideal world the correct target would be the quality of questions and answers.
I always think people calling AI just ‘fancy autocomplete‘ haven’t really tried it yet.
But that's literally what it is. The only reason you can have dialog-like interactions with language models is because they have been trained with special "stop tokens" surrounding dialog, so the model can (generally) auto complete something that looks like a reasons, and then the inference engine can stop producing text when the model produces the stop token.
I cannot remember a time when quora was not broken, so AI certainly didn't ruin it. The UI and their login popups have always been horrible enough that I try to avoid them. But in addition to that, most specific questions - the ones that would benefit the most from a nuanced human answer - get answered with generic copypasta. My impression is that Quota tried gamifying the process of answering... And they succeeded,…
You may have just joined late? Right at the start when there was artificial signup scarcity it was fantastic. Like hn great except for questions