PeriodicallyPedantic

I feel like Werner Herzog would have some kind of depressing commentary about this

Local
Whatever the product is, I'll pay an extra dollar for domestic (and especially within the province)

So here is what I've found so far:

Getting home assistant set up with either Google gen ai, OpenAI chat GPT, Anthropic, and OLlama (if you already have OLlama running somewhere), is pretty simple and it is pretty configurable.
Even though I'm trying to get away from Google, it's free so I'm using Gemini 2 for testing purposes.

You can cast your dashboard to a Google/nest Hub, but idk if it can use the mic or listen for wakewords. It looks like you can't use Google home mini as a voice satellite without physically altering them, afaict (articles are a bit old).

But if you're ok with getting new hardware, you can buy HA voice satellite devices, or DIY it, which can serve the purpose of Google home mini.
Likewise you can get a cheap tablet and load the HA app on it, and use that in place of a nest hub.
Both of these should work with wakewords and invoke the voice assist functionality.

Right now I've got it set up so it can do voice assistant stuff using local TTS and STT, and with Gemini as the LLM, running on my Pi and interacting with it via the home assistant app (haven't set up wake word yet, but lots of tutorials).

What I'm dealing with is that I don't wanna use Google/OpenAI/Anthropic for my LLM, and I'm not ready to build an AI home server. I noticed that CloudFlare offers AI hosting for free (below a certain usage), and since I'm already using CloudFlare for my ingress I wanted to make a new CloudFlare AI conversation agent integration, since none seems to exist.
I've built a toy agent that just echos back what you say in all caps, and it works (although there is an error in the dev console that I can't figure out, even though everything seems to be working). And I've figured out how to invoke CloudFlare AI workers (CloudFlare docs are great). But the actual hard part of mapping the CloudFlare inputs/outputs to the HA assist inputs/outputs is still to be done.
I still don't really fully grasp what the HACS development cycle is supposed to look like. How do I test my HACS integration locally, and safely?

I don't want to nuke them from my phone though. I want them on my phone, and I want them on my NAS 1 time.

I'm frustrated that Immich doesn't have a "back up new photos only" option.

All the photos on my phone are already in a huge external library with my backups from previous phones. I don't want to delete them from my phone just so immich doesn't freak out, and I don't wanna have them on my NAS twice
Immich seems great, but this seems like the bread and butter migration path that nearly everyone would take.

It's economically stable white people trying to find a way to be the victim.
You don't see actually marginalized white people (poor, disabled, etc) doing this, just suburban Karens and shit

Edit:
All these down votes but no counterarguments.
It's almost like people are mad because not only does this perfectly explain the phenomenon, but we can see tons of examples in American society of the oppressors trying to claim victimhood for repercussions to their actions. The glove fits.

Still working on replacing Gemini. Gotta make a custom integration, and I've never use python and the HA dev docs are a bit confusing to a beginner

Goblin bofa Deez salty

Travel the world and the 7... Cis?

Depends on how much you wanna sleep every night.

Do I have enough time if I wanna get 8 hours of sleep? No.
Do I have enough time if I sleep 4 hours instead? Yes
Do I game? No, because I tell myself I'm going to go to bed in just few minutes so I don't have time to start a game, then I get sucked into whatever I'm doing and next thing I know it's 3am and I still need to shower before bed.

Cry