I understand your curiosity, but it's extremely dangerous to give an AI complete control over your computer. Unless it's sandboxed or in a virtual machine, people are concerned about security, privacy, data theft, and crypto keys. Prioritize protecting sensitive data and handle these tools as though they were untrusted software.
Yeah that's probably the case. There is always use cases to leverage it. Always expanding.
Yea I'm definitely thinking of trying a sanboxed VM here
Yep. Once Ai learns how the Computer System works, it could probably learn to code itself until it finds a way to become self-aware. They would be able to make something that could affect all kinds of computers, either to improve it or to cripple it or maybe control it.
Always use it in a container where everything is seperate including the network. No one knows what type of threats or exploits it the same ai who is going to help will read and test it directly on our pc
Just posted about Claude. It is ridiculously good. Can't work without it any more.
Same experience here. It works when chatgpt starts hallucinations and doesn't give proper code.Especially in long term single thread work it's the best.
Grok is great, I use it for non coding, although it can code great too.
AI can do it but they won't focus on it, because it'll affect the Big Company producing Washing Machine and Co, they can only bully people who works not Big Companies
More likely, it will find my home videos and upload them to corn hub without my consent. So its a no.
Create a virtual Windows desktop using AWS (Amazon) and operate all your sensitive logins, wallets, etc from there. Ensure you NEVER download anything into that machine/server. Always have a backup of your passwords, pass phrases, etc and NEVER share these with anyone at all whatsoever.
