• 6 Posts
  • 1.39K Comments
Joined 1 year ago
cake
Cake day: June 29th, 2023

help-circle





  • What are you talking about? I don’t think you understood the concept of decentralised torrent-like hosting.

    I’m currently talking to a peertube hoster about server costs, which I may be able to justify to host my own videos plus a little extra to pitch in for others who can’t justify the expense. Plenty of professional creators could easily justify it as an exit strategy or backup for youtube.

    These conversations are happening, just not with you, presumably because you’re just being negative about it and not actually doing something.


  • Take out the phone part and allow users to host videos in a decentralised way on their home computers and it’s a genuinely good idea though. I have a server running with plenty of storage and reasonable upload speed. I could easily dedicate a terabyte or so, as long as I’m not the sole hoster.

    It would be a hell of a lot cheaper than dedicated hosting. The only issue is legal problems when someone is unknowingly hosting abuse material, which is something that happens from time to time on all services like this, and an individual could be done for distribution without the protection big centralised services have. You’d just have to hope mods are on top of it.

    Actually something like a debrid service but for peertube might work. You can get huge amounts of storage for cheap because a lot of it is shared, you might ask them to host a huge torrent file, but most torrent files serve multiple users, so the cost is distributed. Peertube could work a similar way if it were more mainstream.







  • It’s an illusion. People think that because the language model puts words into sequences like we do, there must be something there. But we know for a fact that it is just word associations. It is fundamentally just predicting the most likely next word and generating it.

    If it helps, we have something akin to an LLM inside our brain, and it does the same limited task. Our brains have distinct centres that do all sorts of recognition and generative tasks, including images, sounds and languge. We’ve made neural networks that do these tasks too, but the difference is that we have a unifying structure that we call “consciousness” that is able to grasp context, and is able to loopback the different centres into one another to achieve all sorts of varied results.

    So we get our internal LLM to sequence words, one word after another, then we loop back those words via the language recognition centre into the context engine, so it can check if the words match the message it intended to create, it checks them against its internal model of the world. If there’s a mismatch, it might ask for different words till it sees the message it wanted to see. This can all be done very fast, and we’re barely aware of it. Or, if it’s feeling lazy today, it might just blurt out the first sentence that sprang to mind and it won’t make sense, and we might call that a brain fart.

    Back in the 80s “automatic writing” took off, which was essentially people tapping into this internal LLM and just letting the words flow out without editing. It was nonesense, but it had this uncanny resemblance to human language, and people thought they were contacting ghosts, because obviously there has to be something there, right? But it’s not, it’s just that it sounds like people.

    These LLMs only produce text forwards, they have no ability to create a sentence, then examine that sentence and see if it matches some internal model of the world. They have no capacity for context. That’s why any question involving A inside B trips them up, because that is fundamentally a question about context. "How many Ws in the sentence “Howard likes strawberries” is a question about context, that’s why they screw it up.

    I don’t think you solve that without creating a real intelligence, because a context engine would necessarily be able to expand its own context arbitrarily. I think allowing an LLM to read its own words back and do some sort of check for fidelity might be one way to bootstrap a context engine into existence, because that check would require it to begin to build an internal model of the world. I suspect the processing power and insights required for that are beyond us for now.








  • Excrubulent@slrpnk.netto196@lemmy.blahaj.zone👮🏻‍♂️ ruleplay
    link
    fedilink
    English
    arrow-up
    3
    arrow-down
    1
    ·
    edit-2
    25 days ago

    There’s nothing wrong with saying clip in this context. The only reason to object is if you’re being annoying and pedantic and you’ve forgotten that there can be more than one term for the same thing.

    Nobody is confused by saying “clip”, nobody thinks you’re charging a glock mag with a clip from the open ejection port. There’s like one gun in history that uses both clips and detachable magazines, so if we’re talking about that specific gun and the difference is operative to the point, then you can complain.

    This is similar to people insisting on “firearm” over “gun”, or “suppressor” over “silencer”, or “round” over “bullet”. Some of them might be more technically rigorous terms but unless you’re discussing the finer points of gunsmithing or ballistics you usually don’t need to care.

    Also, remember the golden rule of pedantry: if you understood well enough to correct them, you don’t have to correct them.