Welcome • User Guide
ToS • Privacy • Canary
Donate • Bugs • License

©2024 Poal.co

1.4K

So, his self-made LLM wrapper has the Linux skill level of a pajeet.

Archive: https://archive.today/XOfeQ

From the post:

>Buck Shlegeris, CEO at Redwood Research, a nonprofit that explores the risks posed by AI, recently learned an amusing but hard lesson in automation when he asked his LLM-powered agent to open a secure connection from his laptop to his desktop machine. "I expected the model would scan the network and find the desktop computer, then stop," Shlegeris explained to The Register via email. "I was surprised that after it found the computer, it decided to continue taking actions, first examining the system and then deciding to do a software update, which it then botched."

So, his self-made LLM wrapper has the Linux skill level of a pajeet. Archive: https://archive.today/XOfeQ From the post: >>Buck Shlegeris, CEO at Redwood Research, a nonprofit that explores the risks posed by AI, recently learned an amusing but hard lesson in automation when he asked his LLM-powered agent to open a secure connection from his laptop to his desktop machine. "I expected the model would scan the network and find the desktop computer, then stop," Shlegeris explained to The Register via email. "I was surprised that after it found the computer, it decided to continue taking actions, first examining the system and then deciding to do a software update, which it then botched."
You are currently inside a comment thread.
Click here to see all the comments (5).
[–] • 1 pt

I was surprised that after it found the computer, it decided to continue taking actions, first examining the system and then deciding to do a software update

Not at all. We are conditioned to believe that keeping software up to date improves security, but this is not universally true. I can see an LLM viewing this as a security requirement based on all the times I read this. At best frequent updates are a treadmill.

No problem no patch is still a valid strategy.

[–] • 0 pt

Just wait until your AI agent votes for you.

[–] • 1 pt

I trained it here. It is voting giant meteor.

[–] • 0 pt

Look at it this way. If it was a ubuntu or similar spin when you log in it tells you if there are updates and even suggests the command(s) to run those updates. It is possible that once it logged into the system it interpreted the text on the screen as more prompts resulting in it attempting to run updates. That is my guess at least.