All of that is worth the extra safety.
480 karma · joined May 31, 2022
A disclaimer is, in my opinion, not enough.
Your position is employers are building the best house with the biggest moat to attract the entire world, with an assumed endless time and budget, and there are no bounds to be set with employees.
In your fictional scenario of unlimited budget and time, sure I grant that an expert should work unguided.
If you’re being paid for your time by someone else, it’s fair to notify them how you plan to use a significant chunk of that money before you do it. Unless of course you were employed to _not_ do that.
I am not suggesting explaining a day or two of work. But it sounds like you’re talking weeks.
There are very few professions we’d consider this acceptable. Implicit in this statement is the assumption that your time-value is only known by you (or free).
You certainly wouldn’t let your contractor improve things on your dime, unplanned, and unexplained. You might even fire them if you discovered they spent the day “refactoring” conduit instead of installing the pot lights you asked them for.
edit: someone pointed out they’ve signed new users on to a US co. 15 months ago. I made the statement without knowing this. they aren’t as capable as I originally claimed.
I’m assuming the attack surface is reduced. I invoke it through a docker container. But this might be a misplaced sense of safety.
Read the post. You might learn something.
What AI have you been using for 5 years of coding?
There’s no proof of your claim and this video is proof of the opposite.
Also, your newspaper is selling the data points it has. If it had more, it would sell more. See: your local paper isn’t selling ads to a car wash six towns over. They do, however, sell ads that align with the political affinities of your local newsrooms area.
Because your parameterized queries have two channels. (1) the query with placeholders, (2) the values to fill in the placeholders. We have nice APIs that hide this fact, but this is indeed how we can escape the second channel without worry.
Your LLM has one channel. The “prompt”. System prompt, user prompt, conversation history, tool calls. All of it is stuffed into the same channel. You can not reliably escape dangerous user input from this single channel.
How?
Does one need to be gullible to believe this? Or will you substantiate your extraordinary claim?
No, you’re paying for “Claude Code” usage.
Yes, Plex _should_ work without an internet gateway. Why? Because it’s a client/server media application; it transcodes media to clients/players over the network.
Plex used to work like this. Actually, it was exclusively unauthenticated. Then early 10s they added optional auth, and eventually allowed you to reserve “server names”, and finally enforced with for running their server. But you can still use a client without auth today. Just read their docs: https://support.plex.tv/articles/200890058-authentication-fo...
You were given helpful advice and a link. I don’t see this being condescending.
And nobody said citizen. Your posts reek of trolling.
You’re contradicting yourself.