• 0 Posts
  • 160 Comments
Joined 1 year ago
Cake day: June 16th, 2023
  • anachronist@midwest.socialtoTechnology@beehaw.org•22 million on blueskyEnglish
    12·
    2 days ago

    I read Chris Webber’s essay and I kinda agree. Bluesky is really just another twitter.

    That being said I think we are entering into an era of diversification, not perhaps how we would like (through federation) but rather, through people understanding finally that the platform itself is making a choice in what kind of content it serves. We used to have this idea that the platform was just a “neutral third party” like a phone company. But in fact, it’s a publisher with its own editorial line. It pushes that line through algorithms and what voices it wants to amplify or suppress.

    As people understand this more, they are going to be much more critical of not just “the media” but also “the platform” and why it chose to show that media to its audience.

  • anachronist@midwest.socialtoTechnology@beehaw.org•Large language models not fit for real-world use, scientists warn — even slight changes cause their world models to collapseEnglish
    31·
    9 days ago

    An important characteristic of a model is “stability.” Stability means that small changes in input produce small changes in output.

    Stability is important for predictability. For instance, suppose you want to make a customer support portal. You add a bot hoping that it will guide the user to the desired workflow. You test the bot by asking it a bunch of variations of questions, probably with some RLHF. But then when it goes to production, people will start asking it variations of questions that you didn’t test (guaranteed). What you want ideally, is that it will map the variants to the best workflow that matches what the customer wants. Second best would be to say “I don’t know.” But what we have are bots who will just generate some crazy off-the-wall crap, and no way to prevent it.

  • anachronist@midwest.socialtoTechnology@lemmy.ml•Mozilla doubling down on ads in FirefoxEnglish
    1·
    2 months ago

    They would have had to build that infrastructure. I’m not saying fundraising is easy. But it’s possible as proven by wikipedia. They could have cut Google loose 10 years ago and said "we’re going to use our runway to try to put together a wikimedia foundation style fundraising operation. I don’t think they can do it now because the trust, goodwill and quite frankly, userbase is gone.

  • anachronist@midwest.socialtoTechnology@lemmy.ml•Mozilla doubling down on ads in FirefoxEnglish
    11·
    2 months ago

    What on earth would that do? The poisonous leadership would not use it to improve the browser nor would they start working for donors instead of Google.

    My point is that there is a funding model that they could have pursued when they still had goodwill and trust. And my hope is if the government finally puts the boot in with Google, then this current version of mozilla will collapse, the rats will leave the ship and hopefully a good browser will emerge the way firefox emerged from netscape.