• 1 Post
  • 66 Comments
Joined 4 years ago
Cake day: May 31st, 2020
  • Ephera@lemmy.mltoOpen Source@lemmy.ml•Ladybird Browser Team Selects Swift as Preferred Language
    81·
    1 month ago

    It isn’t, if you’re already familiar with Rust. That’s all I’m saying. Swift usage is largely isolated to Apple’s ecosystem, which doesn’t have a ton of overlap with the open-source ecosystem.

    And I actually disagree that Rust is overhyped, because it can be used for creating libraries which can be called from virtually any other language, like you can with C and C++. Which means you’re not locked into the Rust/Apple/whatever ecosystem, but instead could be coding the next SQLite without needing to be fluent in footgun.

    From what I can tell, this would theoretically be possible in Swift, but hasn’t been implemented: https://forums.swift.org/t/formalizing-cdecl/40677

    But even if Rust was the most overhyped garbage, it would still be garbage that people are familiar with. ¯\_(ツ)_/¯

  • Ephera@lemmy.mltoTechnology@lemmy.ml•OpenAI could be on the brink of bankruptcy in under 12 months, with projections of $5 billion in losses
    20·
    2 months ago

    I do expect them to receive more funding, but I also expect that to be tied to pricing increases. And I feel like that could break their neck.

    In my team, we’re doing lots of GenAI use-cases and far too often, it’s a matter of slapping a chatbot interface onto a normal SQL database query, just so we can tell our customers and their bosses that we did something with GenAI, because that’s what they’re receiving funding for. Apart from these user interfaces, we’re hardly solving problems with GenAI.

    If the operation costs go up and management starts asking what the pricing for a non-GenAI solution would be like, I expect the answer to be rather devastating for most use-cases.

    Like, there’s maybe still a decent niche in that developing a chatbot interface is likely cheaper than a traditional interface, so maybe new projects might start out with a chatbot interface and later get a regular GUI to reduce operation costs. And of course, there is the niche of actual language processing, for which LLMs are genuinely a good tool. But yeah, going to be interesting how many real-world use-cases remain once the hype dies down.

  • Ephera@lemmy.mltoTechnology@lemmy.ml•Goldman Sachs put out a 31-page-report titled "Gen AI: Too Much Spend, Too Little Benefit?”
    5·
    2 months ago

    I was amazed that we transitioned from one GPU heavy bubble (Crypto) to another (LLM/AI).

    From what I understand, that’s not a coincidence. We’ve had the technology for LLMs for a long time. We just didn’t make them Large, because the training would have taken an eternity.
    Then cryptocurrencies caused the GPU market to boom and then implode, which meant lots of cryptominers were selling their GPUs for dirt cheap. Perfect time to buy a bunch of GPUs, stick them in a server and try to do that eternity-long training.

    And yeah, as for local chatbots, I do also imagine they will continue to be a thing, even if they won’t get any better anymore when the big companies stop doing the training.
    But as you kind of also reasoned, I still don’t expect most of these companies to host their own chatbot…

  • Ephera@lemmy.mltoTechnology@lemmy.ml•Goldman Sachs put out a 31-page-report titled "Gen AI: Too Much Spend, Too Little Benefit?”
    71·
    2 months ago

    Man, I’m so curious when and to what extent this whole hype bubble will implode.
    Like, we’re not talking blockchain, there is some legitimate use-cases for generative AI, which will continue to exist.

    However, there’s also many legitimate use-cases which will not continue to exist, because Microsoft et al are subsidizing GenAI to an insane degree.
    When the hype falls off, investors pull out their money and Microsoft cannot continue subsidizing, which will make prices shoot up for their customers and serve as a rough awakening to all the websites that integrated a crappy chatbot.

    And of course, there’s also the complete fucking bullshit use-cases, which will be gutted immediately when investors/management stop being hyped.

  • Ephera@lemmy.mltoDeGoogle Yourself@lemmy.ml•GBoard replacement?
    0·
    2 months ago

    F-Droid is like a Linux distribution. And basically, the idea behind that is that open-source means nothing in terms of trustworthiness, unless you compile the source code yourself. The source code is what can be audited, whereas the binaries that the developer provides could be filled with malware and you’d never know about it.

    Now, compiling everything yourself is complex and time-consuming, so this is where F-Droid / distros come in. Rather than trusting each and every app developer, you merely need to trust the F-Droid team.

  • Ephera@lemmy.mltoTechnology@lemmy.ml•The Real Reason Why Music Is Getting Worse
    6·
    2 months ago

    In regards to there being an abundance of songs, I’ve also found that this kills off my motivation to make music.

    For one, because well, obviously there’s musicians out there that produce music at a much higher skill level and production value, so it’s easy to just never even try ‘competing’ with them, even though with a bit of effort, I definitely can create something that a certain niche will enjoy.

    But similarly, it also feels like every niche is covered. Any song I’d want to make, I would just need to search long enough and I’d find something similar on the internet.

    I myself don’t have a real demand for the songs I create. I don’t expect to create something that I would find so much better than what everyone else does.
    I do get the bonus of writing exactly the songs that ding the neurons in my brain, but those other musicians get the bonus of having more skill and production value and being the proverbial infinite monkeys with typewriters.

    At this point, I tend to go back and forth between listening to all the excellent music out there, to try to keep my own creativity up to speed, and then for a few weeks, I’ll only listen to relatively mellow songs, so that the songs I’m writing actually sound decent in comparison and I get the motivation to continue working on them.