You are viewing a single comment's thread:
One of our leading AI tools evangalist in my country showed the power of OpenClaw within a week after the original was released; On Television; In a highly viewed talk show. Not many people in the studio seem to be amazed.
Anyways, that wasdn't what I wanted to say.
This guy told all of us to start using OpenClaw on a clean machine without any personal data. Most secure setup. Ok, OpenClaw will not be able to help us on our main machine when doing so, but at least it protects all our personal data by default. I was thinking to buy a new machine, followin ghis advise. Am wondering though if running a VM on my laptop may result in the same, ie run openClaw in a Linux VM, the Linux VM completely silo-ed to the rest of my laptop. VirtualBox comes to mind to use. Do you have any suggestions in this respect?
A VM is "usually" secure. It is isolated from your machine, but there is a small risk of code execution outside of the VM, it's extremely rare and usually is only a threat with very targeted attacks customized for you specifically. In general, it's safe.
Keeping it on a dedicated machine/VM is a very good move. I don't recommend it on your main machine you interact with.
It by nature is very insecure software, and you can lock it down or keep it as free as you want/need.
I have mine deployed in a VM on my Proxmox cluster, so if one fails another takes over in less than 3 minutes.