Mama told me not to come.

She said, that ain’t the way to have fun.

  • 3 Posts
  • 5.51K Comments
Joined 2 years ago
cake
Cake day: June 11th, 2023

help-circle

  • That depends on what you mean by “know.” It generates text from a large bank of hopefully relevant data, and the relevance of the answer depends on how much overlap there is between your query and the data it was trained on. There are different models with different focuses, so pick your model based on what your query is like.

    And yeah, one big issue is the confidence. If users are aware of its limitations, it’s fine, I certainly wouldn’t put my kids in front of one without training them on what it can and can’t be relied on to do. It’s a tool, so users need to know how it’s intended to be used to get value from it.

    My use case is distilling a broad idea into specific things to do a deeper search for, and I use traditional tools for that deeper search. For that it works really well.


  • While true, I think it’s important to note that many buy the Switch for other reasons. My kids wanted a Switch, but I didn’t get it until there were enough games my wife and I really wanted to play. My wife was bummed about Kinect dying and was Ted a replacement for her exercise games, and I had been missing Zelda games, so I got the Switch, some Just Dance games, Ring Fit Adventure, the two Zelda remakes, and a couple games for the kids. The kids have kind of taken it over, but it still fulfills our purposes in getting it.

    My point is that the Switch has a lot more appeal than just shutting kids up for a bit. It’s a good console on its own, and the only console I’m willing to buy. The PS5 and Xbox Series has nothing I’m interested outside of a few exclusives, so my wife and I just play on our PCs and my Steam Deck.


  • My history with consoles is:

    1. Whatever by brother bought
    2. OG Xbox to play Halo
    3. Xbox 360 for Kinect games
    4. Switch - play w/ kids; Smash has been amazing for this
    5. Steam Deck - not a console, but I use it as one; got it to play games in bed

    I play most games on PC because I’m just not as interested in exclusives anymore, except maybe Zelda games, and with BOTW and TOTK, I’m less interested in those (they lost the formula I like).

    I’ll probably get the Switch 2 eventually, but I’ll wait until there’s a game I really want (say, ALttP remake or something), my kids break our OLED Switch, or there’s an OLED Switch 2 with better battery life.


  • I sincerely hope people understand what LLMs are and what they’re aren’t. They’re sophisticated search engines that aggregate results into natural language and refine results based on baked in prompts (in addition to what you provide), and if there are gaps, the LLM invents something to fill it.

    If the model was trained on good data and the baked-in prompt is reasonable, you can get reasonable results. But even in the best case, there’s still the chance that the LLM hallucinates something, that just how they work.

    For most queries, I’m mostly looking for which search terms to use for checking original sources, or sometimes a reference to pull out something I already know, but am having trouble remembering (i.e. I will recognize the correct answer). For those use cases, it’s pretty effective.

    Don’t use an LLM as a source of truth, use it as an aid for finding truth. Be careful out there!




  • I’m not OP and am a dev, but also prefer flat files. Here’s my reasoning:

    • versioning - I use snapshots in my filesystem (BTRFS), which is more than enough, and have a git hosting solution for things I care about more
    • sync is plenty fast on OCIS and Samba, it’s just kinda slow on Nextcloud; I’m sure Seafile is better, but it’s not something I do frequently anyway, especially since backups from devices is automatic and uses a different, fast system
    • incremental - not my use case, most of my files either never change (movies) or are small (text flees)

    My main concerns with Seafile specifically are:

    • developed by a Chinese company and doesn’t seem particularly open to contributions
    • mostly written in C, so there’s a good chance of security vulnerabilities
    • documentation about the disk format isn’t very open, so third party tools don’t really exist
    • main target is larger orgs, so I’m unlikely to get very good support

    With flat files, I can easily switch to a different service if my needs change.


  • Here’s what I’ve used and can recommend:

    • samba - just a network share
    • Nextcloud - full featured cloud suite (calendar, contacts, etc)
    • owncloud infinite scale (OCIS) or the Euro fork [Open Cloud]{https://opencloud.eu/en) - the POSIX driver has a flat file structure and still supports users and shared data; OCIS is designed for larger installations, but running on a smaller, single instance totally works too

    Since you rejected NextCloud, check out the other two. I’m switching from NextCloud to OCIS right now, and I may end up using OpenCloud if development looks stable.




  • I assume you’re talking about Kauffman, who is the founder of LBRY, but that relationship ended when LBRY lost a lawsuit and Odysee was acquired. It is decentralized, using arweave for video hosting and a blockchain for video metadata.

    The main issue w/ Odysee is its near complete lack of moderation, which allows extremists, conspiracy theorists, and other undesirables to earn money. This is because Odysee gives creators the power to moderate their channels, unlike YouTube where most of that is reserved for the platform itself. Odysee is about as free-speech as you get, and that unfortunately allows less desirable content.

    My understanding is that Odysee is essentially what you get if you have P2P (not federated) PeerTube w/ a profit motive.