I have sat in so many meetings looking at shit generated by claude while someone grins and goons in front of it, and I just realise... 99.9% of people in corporate life neither care about the things they produce or present, or can tell a carefully created document from any old shit.
OP is dishonest. The reason for the asymmetry is that OP cuts himself some slack (by way of accepting for himself lower quality work or less strictly reviewed work) for the great benefit it bring him personally, but he doesn't have the same courtesy for his colleagues. He has a higher standard for others than for himself, that's the root source of the bullshit.
AI generated stuff is low quality stuff which you don't accept when it comes from others because you only see the cost. The cost to you of accepting low quality shit, and the benefit to others of being able to claim they delivered something.
But when it comes to you, you don't mind that you're putting out low quality stuff because it's worth the benefit to you being able to claim you delivered something.
People who sound like AI when they write, even if they’re liked and productive, have become identifiable because of the overload of ML text. Maybe at some point we’ll identify precisely what we don’t like about the results of telling AI to write, and we’ll become better for it. Patterns are just neutral, really.
just because a typewriter is awesome for writing an essay on, if I complain that it's crap for banging nails into the wall, doesn't mean that I shouldn't use it for writing essays.
I'm hopeful the annual 13x drop in LLM prices will result in reasonably prized frontier models that can produce something that's actually readable, and that reasons about the code with mental models similar to those of humans, before generating the code. Everyone is in the same mess with their black boxes of unintelligible code, and that's a huge issue for security since transparency and verifiability are the same reason we expect e.g. secure messaging apps to be open source.
But as projects snowball, I wonder if they will ever fit the context window of reasonably prized LLMs.
The real problem isn't readability, it's that my human context is limited and if you throw 1000 pages of readable stuff at me I still cannot get through it in a day unless I throw my LLM at it.
personally, I'm tired of reading 30+ pages long docs with core findings spread across the doc and fit in 1-2 pages. many engineers produce tons of material now, and I spend quite some effort (with AI of course) to make this content make sense
I totally agree that just because everyone gets an app it doesn't mean they like it. Or that because an app is popular it doesn't mean that a majority of people have installed it.
Here's the thing. I'm also willing to believe that people are happy to say they don't like something that they are happy to use. I'm even willing to believe that people have reasons to say they don't like something that they actually do in fact like.
> Here's the thing. I'm also willing to believe that people are happy to say they don't like something that they are happy to use. I'm even willing to believe that people have reasons to say they don't like something that they actually do in fact like.
You seem to be modeling "like" over-simplistically.
Someone can like having ChatGPT give a quick answer to a question they don't really care about, but hate how it has been used to pollute their feed with slop, and hate AI's overall effects on society even while appreciating it in some small ways.
29 comments
[ 3.4 ms ] story [ 46.2 ms ] threadwhere on earth did I say that?
But when it comes to you, you don't mind that you're putting out low quality stuff because it's worth the benefit to you being able to claim you delivered something.
> i f*king detest reading LLM-generated material
Thats quite a dichotomy right there.
just because a typewriter is awesome for writing an essay on, if I complain that it's crap for banging nails into the wall, doesn't mean that I shouldn't use it for writing essays.
https://www.tomshardware.com/tech-industry/artificial-intell...
I'm hopeful the annual 13x drop in LLM prices will result in reasonably prized frontier models that can produce something that's actually readable, and that reasons about the code with mental models similar to those of humans, before generating the code. Everyone is in the same mess with their black boxes of unintelligible code, and that's a huge issue for security since transparency and verifiability are the same reason we expect e.g. secure messaging apps to be open source.
But as projects snowball, I wonder if they will ever fit the context window of reasonably prized LLMs.
https://marketoonist.com/2023/03/ai-written-ai-read.html
Then look at see what the most popular phone app is. Muse. Followed by competitors from Claude, Anthropic, Google , X.
Nobody likes it and everyone wants to use it.
> Then look at see what the most popular phone app is. Muse. Followed by competitors from Claude, Anthropic, Google , X.
> Nobody likes it and everyone wants to use it.
That only has contradiction vibes if you don't think about it too much.
Here's the thing. I'm also willing to believe that people are happy to say they don't like something that they are happy to use. I'm even willing to believe that people have reasons to say they don't like something that they actually do in fact like.
You seem to be modeling "like" over-simplistically.
Someone can like having ChatGPT give a quick answer to a question they don't really care about, but hate how it has been used to pollute their feed with slop, and hate AI's overall effects on society even while appreciating it in some small ways.
Ps: I'm very tired of the all lowercase flex.
"I know I shouldn't buy do or buy X, but in my case it's okay."
P.S. Correctly capitalised, just for you :)