Kay Ohtie

Weird how the chart just prioritizes other corporations and only shows true open-source or self-hosted solutions when there are no non-US options realistically. Hmm.

It does! I have my bedroom one controlled through it and even showing up as a play target for Spotify Connect. I’ve got my speakers I was plugging into my phone to play music before, or into a Raspi briefly, plugged into the 3.5mm jack on that one.

My kitchen one I just leave as-is. I DID modify the ESPHome firmware on each, extending to add an OLED (I think) clock display that also shows remaining time for timers in numbers. I do really like the LED ring animation for timers built-in though, it’s pretty slick!

I ended up picking up two of the Home Assistant Voice PE devices and I’ve been fairly happy with them. I even extended their firmware so I have a clock display on each with one being my bedroom alarm clock even. But even out of the box functionality, as long as you can either run faster-whisper on Home Assistant (or another box), or don’t mind their lighter device-control-only route, is totally solid.

Plus music streaming to them (with an external speaker attached via the 3.5mm jack) is pretty good!

I imagine more as in using them for local voice. Without that, it’s still dependent on connecting HA to Google Home. And outside of a fairly expensive hardware replacement module it ends up being cheaper to go other routes.

There’s a mode for voice control that is even friendly to a Raspi 4 or 5, but it’s very simplistic in control, basically a super lightweight speech to text trained only on device names and aliases. Think the speech to text in late 2000s through early 2010s non-smart phones.

Small models for faster-whisper will run on even my little Dell Micro i5-6500T that I have Home Assistant running on, it’s just a little bit slow, but it absolutely works and is usable speed! I run a larger model currently offloaded to my server, which has an RTX 2070 Super in it, but that’s to make it perform more like how Google used to a long time ago, and it’s unused power most of the time.

They’re trying to make it as accessible as possible for sure. There’s even options to use cloud STT and TTS (they even include it in the Home Assistant Cloud optional feature), but it’s definitely cool as hell to be able to talk to an open-source-design speaker and get a reply and control any switches or lights or even my thermostat and robo vacuum without needing the Internet to work. As long as my Wi-Fi and HA box are up, I’ve got options!

Cookies I can 1000% attest to. The dough gets made quick, sure, but letting it age in the refrigerator for 3-4 days before baking makes a truly excellent cookie.

Varies on the type of course but chocolate chip nearly always wins here. That and with brown butter, also a “take it slow” process.

I don’t think that the point here was trying to do anything to say that the user did anything wrong. I think it’s simply pointing out how frustrating it is that Microsoft’s Insistence on various things, as part of their EEE policy, created this situation to begin with, and that it wouldn’t have even broken if not for that.

I’m pretty sure that the person you replied to was really just lamenting that that this is what broke it. And that fundamentally, Microsoft is getting exactly what they wanted as a result. And it’s just frustrating.

I figured and I know it’s shorthand, it’s my own frustration that said shorthand has partly enabled the anthropomorphism that it’s enjoyed.

Leave the anthropomorphism to pets, plants, and furries, basically. And cars. It’s okay to call cars like that. They know what they did.

I’ve cried at my psychologist place multiple times but each time was not because of her. It was welling up of deep, un-addressed feelings, or a sense of relief of some kind, or sometimes pure catharsis.

Sessions were often exhausting but because of being emotionally engaging and being about things that were deeply emotional to me. I apologized several times for becoming overwhelmed, feeling weird as a chubby grown man breaking down, but she always reassured in a way that made me feel safe, un-judged.

Her goal was always to help me work through last traumas and keep improving living on my own. And she helped tremendously.

There’s definitely some crap therapists out there, and some who are great at some things but terrible with others. Then there’s some who have studied thoroughly and keep up-to-date to help people, because even if it’s taxing for them, their passion is in helping people.

Don’t dismiss horror stories or the fear; dismissal isn’t helpful. Guiding instead with positives, with some good to help ease the fears, is far more helpful.

Great job on something like this! I’ll probably give it a whirl soon, I like Nextcloud but find it clunky sometimes because it’s often a bit more than I need. Maybe breaking it up into Immich + this would help! Thank you for sharing your project!

One thing to note, your comparison against Nextcloud has a partially-incorrect point regarding file upload max size. The client does upload chunking, so is unaffected by the Cloudflare issue as well, but I believe the web client is still affected, just not the apps. https://docs.nextcloud.com/server/latest/admin_manual/configuration_files/big_file_upload_configuration.html#adjust-chunk-size-on-nextcloud-side

I suspect a few others may be as well, but I’m only familiar with the Nextcloud one because that’s what I’ve been running, and discovered in making sure I could still upload video files recorded while out and about.

Also love that it looks like a simpler install!

As someone living in a red state, every city here is referred to as a “blue city”, whether people are moving to it or not. State color has shit all to do with it and tends to be how much did we make the assumption land votes.

whether it’s telling the truth

“whether the output is correct or a mishmash”

“Truth” implies understanding that these don’t have, and because of the underlying method the models use to generate plausible-looking responses based on training data, there is no “truth” or “lying” because they don’t actually “know” any of it.

I know this comes off probably as super pedantic, and it definitely is at least a little pedantic, but the anthropomorphism shown towards these things is half the reason they’re trusted.

That and how much ChatGPT flatters people.

That’s how you can tell if someone is into latex (kink), they don’t feel comfortable calling LaTeX (tech) by the same pronunciation around people.

Python is compiled at “runtime” to a similar OS+arch byte-code minus ELF headers that Linux binaries typically have from gcc.

My point was it’s a stupid distinction and worthless when the other points about poor implementations of common language frameworks are plenty on their own is all, and it’s needlessly snobbish.

As far as class variable reference however I wish more languages self-referenced. In my eyes it makes it far clearer at a given line of code glance as to where the hell a value came from as opposed to just by name. I feel a keyword like self::variableName, or maybe more aptly &self as a pointer to reference in C++ would be very clear, like Rust does, which is very much, by the original definition, a programming language instead of scripting. Even Java, which is definitely not a scripting language though is still run inside a virtual machine, uses this. I don’t personally like the term versus self, but eh.

Though if you want a hammer in a screw-driven world look no further than Electron. I think it puts anyone else’s even purposeful attempts at such to shame.

C/C++ are just scripting languages that have to become OS+arch -specific byte-code before execution.

Wait there’s a “that”???

Same. Not married, but my partner always had to put in earplugs when I slept with them because I snored so loud, and they were always silent with their CPAP, which was oddly soothing of a sound. Finally got diagnosed after my doctor AND partner both pressured me to. I sleep fairly soundly most nights now. Even got a new mask type yesterday, a minimal contact full face, and I love it. Just gotta adjust my mustache maintenance routine which…to be fair it was already in dire need of, being bushy and rough and uncomfortable, and that upper edge pushing the hairs right back into my face made that all the more obvious. Oof.

I love the feeling of neurons rewiring to form a new pathway of understanding. Or whatever the hell it is. At 38, it’s a pleasure finding I can still learn and build new skills.

Playing Beat Saber and hitting a plateau only to find my focus starts to evaporate over the course of a hard track as I find that flow, that path to just being in it, each skill plateau merely being temporary, is great. Playing guitar and slowly starting to wire my brain for the pathway for barre chords and faster movement along the frets is a crazy feeling. That sense of finally finding the pathways for singing to operate even SLIGHTLY separately from the rhythm of the guitar, those glimpses of polyrhythm? Addicting.

If you’re able, I hope you can teach him to find that pleasure of not mastery, but evolving strengths. Maybe it’s like an RPG where skills can be leveled up over time the more you use them. I know all too well the frustration of imperfection to start, ADHD during the 90s and the whole “perfect student” pressure created a lot I had to undo and still am, but each time I can break free of that it’s rewarding.

It’s processing data alright, it processes the atomic and cellular structures of grass and fingers into spinach and flesh paste.

And likewise, neither it, nor any LLM, are making decisions at all.

Is a plinko disc making decisions as it tumbles from the top to the bottom through all those pegs? Is the board making the decision? Or is it neither and simply mathematics plus random chance being roped in for randomness? That is exactly what LLMs do.

Terms like “decision” and “lie” and “know” are all things that just do not apply to an LLM, just like your phone keyboard doesn’t know what the fuck “what” and “the” are, it just has a lookup table that includes how “what” is often followed by “is” and “the”, and “the” is frequently followed by “fuck”. But it doesn’t “know” that in any meaning of the word “know”.

This is what we mean when we say not to personify. A training set of data, even factual, just is converted into a series of matrices of vectors that include those patterns, but not the information itself. “Sky is blue” is not something you can grep from the resulting blob, nor the hex equivalent, or anything else. It simply contains indexed patterns that map those arrangements of letters, over and over.

So yes, they’re doing what they’re programmed to do precisely. It’s just that “what they’re programmed to do” is only “mimic patterns of word arrangements”, and not “know facts”. These things work at a far lower level than that concept.

By this logic, a lawnmower “thinks” my fingers are grass.