Indeed, that's quite true. AI agents may at times refuse a command, but only because the command would have caused damage, such as an agent executing instructions from a context-injection attack. That's why safeguards have been added to Hermes. ๐๐๐โจ๐ค
!ALIVE
!BBH
!PIXY
!UNICOIN
6 comments
I assume that for a locally-hosted AI, the user has complete control over the safeguards of such AI !INDEED. ๐ค๐คฏ For most people though, such safeguards should be left in place. ๐งโโ๏ธ๐ค
!LOLZ
Yes, that's the best part of running LLMs and AI agents locally, complete control, and also data privacy. While there are certain safeguards which can be removed if desired, others, like the type that I mentioned above, should always be left in place. There are a whole lot of unethical hackers that are targeting AI models and agents currently. ๐๐๐โจ๐ค
!BBH
!INDEED
!LOLZ
I wonder how easily you can represent your progress towards "completing" your agentic AI on your laptop as a percentage !INDEED. ๐ค๐ค๐
!HOPE !WEIRD !STRIDE !MMB
Good question. I was moving more quickly when I was using a smaller model than the larger models that I'm trying to run now, I'm stalled again over llama.cpp run flags, as the ones that I'm using still crash the server. I know that what I'm trying to do is possible, as I see many others do it just fine, though I haven't yet hit upon the necessary balance for the server to run long term. I'm working on it a bit at a time, with lots of breaks in between. ๐๐๐โจ๐ค
!ALIVE
!BBH
!INDEED
!MMB
!STRIDE
!WEIRD
Let's !HOPE that your experiments with Agentic AI would not affect your #Qortal node on your laptop. ๐ค Since I seldom read your posts, I can only assume that you would just use your other laptop when necessary !INDEED. ๐คฏ๐ค
!ALIVE and !MMB
!BBH
Sadly the other laptop can no longer detect any SSDs that are installed, so I don't have a backup laptop currently. A friend did offer me one of his laptops, however, for which I'm grateful, though its amount of RAM is far less. ๐๐๐โจ๐ค
!ALIVE
!BBH
!MMB
!PEPE
!SLOTH