+1 for Proxmox, has been a fun experience as there are plenty of resources and helper scripts to get you off the ground, jellyfin was the first thing I migrated from my PC, hardware encoding may give you a bit of a tussle but nothing unsolveable. Also note Proxmox is Debian under the hood, so you may find it easy to work with. I looked into unraid, it seems great if all you’re doing for the most part is storage, if you want Linux containers and virtual machines, proxmox js your bet.

I got a small 4 bay 2U server from a friend on the cheap, 1000$ should get you relatively nice new or slightly older used hardware. Even just a PC with a nice amount of drive bays will get you started. And drives are cheap, a raid 1 setup was one of the things I did.

In the end I’ll likely get a separate NAS rack server just to segregate functions, but as of now I simply have a Proxmox LXC mounted to my NAS drives and runs samba to expose them.

Tailscale is a nice set and forget solution for VPN access, I ended up going the route of getting an SSL certified domain and beefing up my firewall a bit. The bit I’ve messed with it it certainly has a learning curve greater than openvpn, but is much more hardened and versatile.

As for pihole, I’ve found AdGuard Home to be just about a suitable replacement, and can be installed along openwrt, though I have a bit of an unconventional router with 512MB of RAM so YMMV

I got a 1U 32GB and a 4 core Xeon for 20$, you just have to make at least 3 more posts

We already do! And on the cheap! I have a Coral TPU running for presence detection on some security cameras, I’m pretty sure they can run LLMs but I haven’t looked around.

GPT4ALL runs rather well on a 2060 and I would only imagine a lot better on newer hardware

Now THAT is the AI innovation I’m here for

We do not have a rigorous model of the brain, yet we have designed LLMs. Experts of decades in ML recognize that there is no intelligence happening here, because yes, we don’t understand intelligence, certainly not enough to build one.

If we want to take from definitions, here is Merriam Webster

(1)

: the ability to learn or understand or to deal with new or trying >situations : reason

also : the skilled use of reason

(2)

: the ability to apply knowledge to manipulate one’s >environment or to think abstractly as measured by objective >criteria (such as tests)

The context stack is the closest thing we have to being able to retain and apply old info to newer context, the rest is in the name. Generative Pre-Trained language models, their given output is baked by a statiscial model finding similar text, also coined Stocastic parrots by some ML researchers, I find it to be a more fitting name. There’s also no doubt of their potential (and already practiced) utility, but a long shot of being able to be considered a person by law.

I don’t want to spam this link but seriously watch this 3blue1brown video on how text transformers work. You’re right on that last part, but its a far fetch from an intelligence. Just a very intelligent use of statistical methods. But its precisely that reason that reason it can be “convinced”, because parameters restraining its output have to be weighed into the model, so its just a statistic that will fail.

Im not intending to downplay the significance of GPTs, but we need to baseline the hype around them before we can discuss where AI goes next, and what it can mean for people. Also far before we use it for any secure services, because we’ve already seen what can happen

Building my own training set is something I would certainly want to do eventually. Ive been messing with Mistral Instruct using GPT4ALL and its genuinely impressive how quick my 2060 can hallucinate relatively accurate information, but its also evident of limitations. IE I tell it I do not want to use AWS or another cloud hosting service, it will just return a list of suggested services not including AWS. Most certainly a limit of its training data but still impressive.

Anyone suggesting to use LLMs to manage people or resources are better off flipping a coin on every thought, more than likely companies who are insistent on it will go belly up soon enough

The fallout of image generation will be even more incredible imo. Even if models do become even more capable, training off of post-'21 data will become increasingly polluted and difficult to distinguish as models improve their output, which inevitably leads to model collapse. At least until we have a standardized way of flagging generated images opposed to real ones, but I don’t really like that future.

Just on a tangent, openai claiming video models will help “AGI” understand the world around it is laughable to me. 3blue1brown released a very informative video on how text transformers work, and in principal all “AI” is at the moment is very clever statistics and lots of matrix multiplication. How our minds process and retain information is by far more complicated, as we don’t fully understand ourselves yet and we are a grand leap away from ever emulating a true mind.

All that to say is I can’t wait for people to realize: oh hey that is just to try to replace talent in film production coming from silicon valley

What provider? Mine off of porkbun was 30$ for 3 years at 14$ a year after discount and I managed to snag a relatively personal & unique name

ProtonDB comments… at the very least say if you’re using an nvidia or AMD card ffs

I was attempting this, but TP link doesnt actually care to tag their different SSIDs to vlans and don’t provide the configuration to, I only found that their guest may be tagged on some models. Just a word of caution, I think I’ll have to use IP range filters to achieve this

I was genuinely concerened I had a skill issue with NC, glad I’m not alone

Its so weirdly addictive, started with proxmox, then to home assistant, now I have frigate handling PoE cameras, every bedroom has a morning alarm light automation, the vacuum starts itself, its ridiculous

Hey don’t mean to hijack, I know tailscale isnt necessarily a tunnel (correct me if I’m wrong) but does it provide the typical issues with apps like android auto not working while connected to a VPN?

Thank you very much for that final note, I just love misleading marketing tactics and actually may not have noticed. Cheers!

Issue is assigning static IPs to devices like nest cameras and the dubiously branded vacuum cleaner, NAT alone should be enough to restrict connection even if the adapter its on is configured as lan then? Im a little in over my head as far as routing goes, wanting to self host some web server and I want to take as many precautions as necessary (maybe even overkill)

What if I like grubbing around? What if I like when updates give me hell?

shudder

Was an Asus laptop, I’m yet to struggle so much again

Out of pocket I would assume nonfree packages and hardware shenanigans: binary blobs, etc. …

The latter imo is only really a concern if you’re being targeted hydrate actors, and you got bigger problems then

The best way to ensure your data lasts a long time is to use a laser to beam it to the darkest part of the sky. Read speed is abysmal though