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.
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.
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.
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
> yeah, ask people what they think of LLM's/AI and a majority are split between either disliking it or fearing it or maybe both.
> 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. Even accepting your all statements as facts, none of them support your conclusion without serious misunderstandings.
Like isn't it obvious that a person downloading a phone app doesn't mean they like it? Or that a high ranking on an app store doesn't mean even a large plurality of people have downloaded it?
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.
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.
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.
A lot of internal presentations and documents are never looked at or read; they're basically used as proof of work.
eating_cake && having_cake == false
With news like
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.
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.
Why are you censoring yourself in your own blog?
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.
My thoughts on AI: https://nonograph.com/some-disorganized-thoughts-about-artif...
AI written, AI read
https://marketoonist.com/2023/03/ai-written-ai-read.html
Everyone finds their own use case acceptable. It's the way others use it that's terrible.
Ps: I'm very tired of the all lowercase flex.
You can apply this logic to most of people's behavior in capitalism.
"I know I shouldn't buy do or buy X, but in my case it's okay."
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
Tools. They're all about how you use them.
Honestly? How about, They're all about how _you_ use them.
yeah, ask people what they think of LLM's/AI and a majority are split between either disliking it or fearing it or maybe both.
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.
> yeah, ask people what they think of LLM's/AI and a majority are split between either disliking it or fearing it or maybe both.
> 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. Even accepting your all statements as facts, none of them support your conclusion without serious misunderstandings.
Like isn't it obvious that a person downloading a phone app doesn't mean they like it? Or that a high ranking on an app store doesn't mean even a large plurality of people have downloaded it?
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.
[dead]
"Please, could you review this pull request?"
The PR description alone, I counted it, was 25k words.> i f*king love LLMs for generating code
> i f*king detest reading LLM-generated material
Thats quite a dichotomy right there.
It’s a classic moment, feel the pain others cause you and miss the pain you cause others.
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.
What high standards?
[dead]