If you have some spare tokens:
- check security of your home network, of apps you use, custom dev software. Use models that don't complain
- write documentation for your home setup and keep it up to...
Agree with most of it.
But you don't need AI in the majority of the the cases. If you start using Claude on your home network you're spilling the beans on your private infrastructure to Antrophic.
Go at it the hard way, and learn something in the process. Also:
- only keep apps online that you actually use
- firewall outgoing traffic in a least privilege fashion
- set up an isolated guest network for devices from friends/visitors
Before going at it understand the threat model; a home user is mostly at risk by downloading malware/spyware from untrusted sources. This can be almost completely mitigated by firewalling and disabling JavaScript in your browser by default.
A socks or http proxy to allow connectivity on an per app basis can also work very well.
But you don't need AI in the majority of the the cases. If you start using Claude on your home network you're spilling the beans on your private infrastructure to Antrophic.
Go at it the hard way, and learn something in the process. Also:
- only keep apps online that you actually use
- firewall outgoing traffic in a least privilege fashion
- set up an isolated guest network for devices from friends/visitors
Before going at it understand the threat model; a home user is mostly at risk by downloading malware/spyware from untrusted sources. This can be almost completely mitigated by firewalling and disabling JavaScript in your browser by default.
A socks or http proxy to allow connectivity on an per app basis can also work very well.
1❤️1
