I usually connect with my server via ssh in a terminal and run basic commands. What’s a better, more efficient and modern way of doing that? Especially considering ai and documentation along the way? I wonder if there’s a better approach than “connect from remote and act local”. Is there a method to “code local and push to remote”?
I use a fedora server with podman, caddyfile and vi.
I looked that up and am astonished that you let an agent range free on your server?!
I bet you I do less work and have a more efficient and locked down network, file permissions, fully rootless, updated on the hour, 100% in code, and observable setup than 98% of you just running sudo apt upgrade and praying. But yeah, my little agent working in an nemoclaw sandbox is going to go rogue and destroy my server somehow and I definitely believe the AI hype bros that are definitely “accidentally “ are hacking orgs and not going to end the human race🙄
Man, the ignorance and prejudice over how llm work is insane.
It’s not about “going rogue”. It’s about making mistakes. It’s not a hypothetical. It has happened many times to people who were as confident as you seem to be with their setup. There’s a proverb in my language that roughly translates to “some learn by looking at others, while some need to experience it by themselves”. Maybe you’re in the latter category.
Humans definitely don’t make mistakes. Especially not with their jank home servers. All of it done without shortcuts and auditable in code, with a standard process.
You’ve set some imaginary high bar that is rarely actually met outside of nerds shit talking on the internet and a small percentage of users.
“I can make mistakes, so I’ll rather use a slop machine that’s known to make mistakes.”
Great strategy, stranger.
Oops, slop machine broke something.
“Hey slop machine, you fucked up. Restore the last backup and document the mistake you made in AGENTS.md so we avoid making that mistake again. Im going to go make a coffee, brb”
“Got it, deleting your home directory.”
So again, because we make backups before letting AI touch anything:
“Hey slop machine, you fucked up. Restore the last backup and document the mistake you made in AGENTS.md so we avoid making that mistake again. Im going to go make a coffee, brb”
Sigh…