Interesting... I ran across stuff about Razer's Project Ava ( https://www.razer.com/razer-ava ) an AI assistant program that interfaces with the user via what looks like a cartoon woman (or man, the character I gather can be selected by the user) in a giant test tube. The system does voice recognition and voice response (in a voice appropriate for the gender of the cartoon character selected). Seriously cute, but do I trust Razer (or any other company) to run an AI in my apartment? @#$% NO!!! Still, I have been thinking, I have some Raspberry PI 7" LCD screens, some old (low resolution) webcams and some old Raspberry PI computers gathering dust. I also do have a 3D printer, so some small custom plastic parts are not an issue for me. So, the questions I currently have are : - Is this something worth doing? The one use case I can currently think of is a front end for Home Assistant (which turns lights on / off in my apartment and monitors some environmental sensors). - Is this something I could do FULLY local? I don't want anything that depends on the Internet to keep working. I also don't want to depend on a PC running Microsoft Windows for several reasons. - Is this something I can do on a next to nothing budget? As in if I have to spend say $20 for an extra speaker, and/or a few dollars worth of 3D printer plastic no problem. If I have to buy a new 16GB RAM Raspberry Pi 5 with an AI add-on card given current (insane) RAM prices, that is an issue. In other words can I do this (largely) with stuff I already have? This guy comes up with something in the ballpark of what I would want, but ... I don't want the costs of custom PCBs (printed circuit boards) or what I regard as an oddball SBC (single board computer). Still, interesting : https://youtu.be/WgqDaqqto90?si=6ulXExEu_hXGMTAf