• 80 Posts
  • 353 Comments
Joined 3 years ago
cake
Cake day: July 9th, 2023

help-circle
  • For my trips I use Trek, AirTrail, and Dawarich.

    I picked Trek because it was most similar to Wanderlog which is what I used to use and it supported SSO. It seems to have a strong community and feature delivery, though my challenges with it is that it feels slow to load and its somewhat complicated to navigate.

    Trek has a beta Android and iOS app.

    Then AirTrail to track flights and see a cool display of everywhere I’ve flown and Dawarich shows my location history. Trek can then pull in my Dawarich and AirTrail history.














  • WebPKI (the term for certificates used for web browser as opposed to the general usage of certificates outside of the web) requires that all certificates now be included in a public, widely used certificate transparency log. That fact gets included in the certificate delivered to your browser. Your browser will reject any certificate* that was not included in a certificate transparency log.

    Take your instance discuss.tchncs.de for example, we can see all the certificates here. If you looked at that list, do you see anything wrong? Probably not.

    But you’re on to something. A malicious certificate is minted and it’s in the CT log, so what? I don’t check the CT log every time I go to a new webpage. Instead, CT logs are more able giving information to security organizations and companies to detect violations under their own properties. A company like amazon or google would setup CT log monitors to continually monitor for certificates issued for their domains. If they detect one, and they didn’t request it, they can flag it to the cert authority board and it can get CAs pulled from the root trust stores. I can (and probably should find time to) setup certspotter on my own domains.



  • It means they release the model so you can run it yourself with enough hardware. Kind of like open source. Governments and large companies should be self-hosting this kind of stuff because of security, availability, and IP-leak risks anyway.

    With LLMs, there is some risk of hidden malicious actions embedded in the weights that you can see until you ask it a magic question and it says “activate skynet.”

    Whether or not the government is doing this because of fears of hidden actions in the weights or just because OpenAI/Anthropic CEOs told the government to give them more money is unclear.


  • Yes, I run Ollama on a NVIDIA 1080 Ti at home. I run granite4:3b for my Home Assistant assistant for my voice assistant, paperless-ngx, OpenWebUI for chat, and a few other things. For speech to text, I run a custom-built hybrid speech-to-phrase + parakeetTTS (requests that match speech-to-phrase take priority over parakeet.) I run GOModel to route requests to my local Ollama instance, or DeepInfra (cloud-based for larger models), or Tinfoil (cloud-based for encrypted LLM chats.)

    As much as I love to use local-everything, only my smarthome+basic chats go local. Trying to fit both the STT and LLM models on the same GPU at the same time while also using this computer as a media console is challenging. I’ve recently started looking at the Hailo m.2 accelerator board to see if I can shift the STT to it.


  • That doesn’t check out when considering model size. Unless that’s what you’re trying to say that self-hosted models force people to use smaller models. Just from an operations perspective a data center had economies of scale and the ability to batch up multiple independent user’s requests which should reduce the per token kWh along with the fact that your utilization will be higher vs somebody’s GPU at home would be idling.

    That said I do believe there are negative environmental impacts from AI like always using giant frontier models for everything instead of smaller fine tuned models.

    EDIT: I see a bunch of replies on the other comment, which all raise valid points, but that’s definitely not the way most readers are going to interpret the original post and moves the goal posts.