___
These micro-examples are a reminder that corruption is a part of every human system, no matter how perfect the design.
There will always be concertgoers cutting the unwatched fence to sneak in for free.
The only plausible solution is elective transparency. Either your company and financial metadata are available for independent third party review, and records retained as defined, or else you’re not a company.
Don’t ascribe to it, get boycotted.
These mini pcs are a godsend: 12th Gen Intel N100 Firewall Computer Soft Router 4x 2.5G i226 i225 LAN NVMe Industrial Fanless Mini PC https://a.aliexpress.com/_mPBQflM
Cheap and more than powerful for opnsense and pi-hole on top.
Home assistant just run the new install script on proxmox to setup a vm. Copy your data over, usb pass through your radio hw and good to go.
Docker on lxc also works and you can probably swarm your containers over.
TP-Link Omada EAP line is rock solid for cheap managed access points. You can even run the controller in an lxc full local.
It won’t murder you if that’s what you’re asking. You can always monitor your egress traffic if you’re concerned. Ultimately, for home use it likely doesn’t matter.
Spying today takes the form of custom silicon branded as the original with extra transistors added before fabbing. Usually higher nm processes since they can’t fab on the latest nodes.
Knowing this, the chance of getting a counterfeit chip on a name brand switch is non-negligible as it is, and those parts come from China too.
The firmware might be a concern, but again, you need to be a valuable target.
Most people here don’t understand what this is saying.
We’ve had “pure” human generated data, verifiably so since LLMs and ImageGen didn’t exist. Any bot generated data was easily filterable due to lack of sophistication.
ChatGPT and SD3 enter the chat, generate nearly indistinguishable data from humans, but with a few errors here and there. These errors while few, are spectacular and make no sense to the training data.
2 years later, the internet is saturated with generated content. The old datasets are like gold now, since none of the new data is verifiably human.
This matters when you’ve played with local machine learning and understand how these machines “think”. If you feed an AI generated set to an AI as training data, it learns the mistakes as well as the data. Every generation it’s like mutations form until eventually it just produces garbage.
Training models on generated sets slowly by surely fail without a human touch. Scale this concept to the net fractionally. When 50% of your dataset is machine generated, 50% of your new model trained on it will begin to deteriorate. Do this long enough and that 50% becomes 60 to 70 and beyond.
Human creativity and thought have yet to be replicated. These models have no human ability to be discerning or sleep to recover errors. They simply learn imperfectly and generate new less perfect data in a digestible form.
I think you need to stop looking at yourself.
I blame developers who use discord. The interface is a mess, the company is enshitifying, and you can’t archive the past.
It maybe easier, but I question why anyone would willingly choose it over most anything else.
Note that this is different from being a dick about it and harassing people.
List off the apps you use, find alternatives or figure out what you can run in wine if necessary.
Learn a little about your distro’s file structure and differences. It can save you hours troubleshooting.
Stay away from nvidia cards where possible, nothing but problems till I moved to AMD.
