From fundraising more than $100 million to hemorrhaging top talent the firm that makes Stable Diffusion has had a heck of a year.
  • AutoTL;DR@lemmings.worldBEnglish
    642·
    1 year ago

    This is the best summary I could come up with:

    From fundraising more than $100 million at the end of 2022 to hemorrhaging top talent by mid-2023, Stability AI — the firm that funds and supports development of the open source Stable Diffusion image generator — has had a helluva year.

    Though Mostaque insists that “churn” is a common practice in startups while trying to establish “cultural fit” between employee and company, interviews with several former and current people involved with the project say the CEO’s lofty vision often doesn’t match up to the reality of his day-to-day ability as a business leader.

    Earlier this summer, Forbes published an exposé that highlighted his “history of exaggeration,” and in its opening lines notes that Mostaque’s claim that he has a master’s degree from Oxford didn’t hold up to scrutiny.

    Case in point: in a statement to Futurism, a Stability representative said that because the company “is in the innovation business,” it is “well aware that any time a new path is taken in any field, there will be critics and skeptics.”

    “That reality is no different for the field of generative AI, which has taken the world by storm because of its potential to be the greatest disruptor of our time,” the statement continued.

    We remain focused on developing the best open language and image models for millions of users worldwide, and our work is just beginning."

    I’m a bot and I’m open source!

  • int-argc-char-argv@kbin.social
    29·
    1 year ago

    Emad is a lot like how I imagine Elon without the money.

    The more I read, the more it seems Runway, not Stability, deserved to release what would eventually become Stable Diffusion.

    • dingleberry@discuss.tchncs.de
      8·
      1 year ago

      All you need to know about him is he spends all his waking hours approaching different publications and asking them to interview him for his thoughts on AI. He is from the new batch of muskulencers.

  • Rumblestiltskin@lemmy.ca
    251·
    1 year ago

    Seems to be just a hit piece. I only care if it does cool things and it’s open source. None of the investor squabbling matter to me.

  • Frank [he/him, he/him]@hexbear.netEnglish
    241·
    1 year ago

    Techbros gonna techbro. This barely counts as gossip. I wish them all a very accidentally trip in to the torment nexus where they suffer for a simulated eternity or something.

  • Send_me_nude_girls@feddit.de
    9·
    1 year ago

    Sad to read. I hope they get their shit together. SD is the only option if you don’t want to subscribe to any censored alternative.

  • andruid@lemmy.ml
    81·
    1 year ago

    Hopefully countries looking for data sovereignty but also want to use generative AI start looking to using them for this before the company dries up and proprietary AI running only in US data centers become the state of the art and defacto place to go.

    I mean, how long has it taken for cloud offerings to start to catch up to AWS.

    • QHC@lemmy.world
      261·
      1 year ago

      One of the most prominent generative AI tools that has been making headlines in the last few months. Stable Diffusion is specifically meant to create fully artificial, photo-realistic images. If you’ve ever seen one of those “all of these people don’t exist” montages, it was almost certainly using images generated by this tool.

          • BaconIsAVeg@lemmy.mlEnglish
            1·
            1 year ago

            These were deities for a homebrew campaign, and the DM had already provided their domain, element, and symbol (i.e. war, fire, stallion). I usually just generate 4 images at a time (only takes a few seconds on a 3090) and pick the one I like the best. Sometimes I’ll generate 2-3 sets of 4, but not often if I don’t have a clear idea of exactly what I’m looking for.

            If it’s something really specific I need, I could spend hours using in painting and various noise/models to get what I want.

            Edit: oops, I was thinking of a different montage I did recently: https://pixelfed.social/i/web/post/595611323719481231

            The previous linked image was much the same process, though the prompts were more detailed as the other players had provided more information on their character’s appearance.

    • JJROKCZ@lemmy.worldEnglish
      115·
      1 year ago

      Company makes a tool that generates “ai” images. Mostly used by people that want to make porn that looks like real people but is not real people.

      • ThunderingJerboa@kbin.social
        181·
        1 year ago

        Which in a weird way is an ethical way to make porn with no “exploitation” involved with the process and there will always be a market for the “real” thing for the market. Now the AI art people can be incredibly thirsty but I do feel it is a bit reductive to say its mostly only for porn when the applications of it can be quite useful especially for hobbyists (who I feel aren’t profiting solely on those creations). I see great use for generative “art” for tabletop in terms of character sheets/custom monsters/settings/ and even maps. AI art isn’t necessarily a bad thing its just the fear of what large corporations will do with it that is really the scary part (also bullshit scammers who steal X artists work and try to resell it as that artist)

    • hempster@lemm.ee
      3·
      1 year ago

      They ushered “AI” even before ChatGPT. It was “The” rad back in August 2022.