from Hacker News

Leaked deck raises questions over Stability AI’s Series A pitch to investors

by mjaques on 4/23/23, 8:40 PM with 66 comments

  • by capableweb on 4/23/23, 9:32 PM

    The model was initially released in August 22, 2022, and everything from the GitHub organization that hosted the code, to the actual licenses being around and all the communication made it very clear that there was a bunch of groups involved with the model; Runway, CompVis, and Stability AI

    Then Stability AI raised money from investors in October 2022.

    Are you really telling me these investors didn't even do any sort of due diligence (not even "barely any" but literally any at all), didn't realized that Stable Diffusion was the work of many, and checked the license of the model and code?

    That's a bit far fetched, to be honest.

    Supposedly, this "leaked pitch deck" is supposed to show us that they lied to investors, but where is the pitch deck itself? Seemingly, only one of the slides is in the article, how could anyone reach any sort of conclusion based on just one slide?

    Edit: The article seems to be some sort of clickbait trash that is so rampant around the web... One selected part:

    > The Stable Diffusion code was released by researchers at LMU Munich in April 2022.

    Links to https://arxiv.org/abs/2112.10752 which is the Version 2 of the paper, indeed released in April 2022. However, that paper is about, and links to https://github.com/CompVis/latent-diffusion which is not Stable Diffusion, it's Latent Diffusion. Stable Diffusion was released in August 2022, and is at https://github.com/CompVis/stable-diffusion

  • by amrrs on 4/23/23, 9:01 PM

    I don't remember Stability claiming in any public forum that they solely developed Stable Diffusion. In fact the first stable diffusion model on Hugging Face is under CompVis Org.

    They had an issue with Runway with a model takedown or something but I think that was sorted out later on.

    I don't know why the article seems desparately trying to show Stability in bad light while they have in fact funded a lot of AI model development and helped release it with permissive licence

  • by ericflo on 4/23/23, 9:07 PM

    I think in the world of ML, contributing that much compute should count as co-creation. There is a lot of code published by academia that just needs compute and data, but they don't receive it. Stability deserves credit for Doing the Thing. IP rights are another thing but that's a whole subtree of legal questions that society is barreling towards.

  • by KaoruAoiShiho on 4/23/23, 9:31 PM

    Dumb hit-piece, Stability's contribution was vital to the creation of those models. Article doesn't even make the claim that the investors were confused on if Stability owned the models, it just insinuates they were for absolutely no reason.

  • by sillysaurusx on 4/23/23, 9:41 PM

    Whoa. So I got a DM on Twitter about a week ago thanking me for pushing back on Emad’s claims about cluster pricing. I didn’t really take it seriously (or understand it) till now:

    https://i.imgur.com/5XAndrR.jpg

    https://i.imgur.com/OAZeZEl.jpg

    So for whatever that’s worth, here’s a claim that predates this article by a week that emad was lying about stability origins to investors. I’ve trimmed the name off the DM for obvious reasons, but it’s from someone who seems credible (though I don’t know them).

    I don’t like posting random unsubstantiated accusations without context. But this article just provided both.

    As for what he lied about (or stretched the truth), I don’t know. I don’t even know if it matters. But this is why I never lie to investors, and why I’ve always really felt negatively towards colleagues who stretch the truth. It’s all kinds of misleading, and it’s a kind of misleading that tends to matter.

    It would also make me upset as an investor if I found out that someone had fooled me. So even if it’d be to my advantage to lie for money, I can’t imagine doing it. I’d rather live in relative obscurity.

  • by nostromo on 4/23/23, 9:23 PM

    Lying and fraud aren’t ok, and founders should be held to account for that.

    But this doesn’t seem like that. It seems like any disappointed investors just didn’t do the bare minimum you’d expect with an investment of $100m.

  • by exizt88 on 4/23/23, 9:35 PM

    Another poorly reported hit-piece from Sifted. This is a non-event. Anyone who’s been close to fundraising knows that investors expect founders to build a narrative and do their own due diligence.

  • by neximo64 on 4/23/23, 9:26 PM

    What is the story here? Stability funded a model that came off the LMU paper?

    They certainly did that better than the way OpenAI did, or Midjourney did.

  • by napier on 4/23/23, 9:48 PM

    What a clumsy half assed hack attempt at a hatchet job.

  • by yawnxyz on 4/23/23, 10:04 PM

    They should have interviewed the investors to understand where they saw the opportunity.

    I think an ecosystem of users coming back to perform (and pay for) all kinds of generative AI tasks is totally worth the investment.

    I can't figure out how Stability hasn't been able to make revenue though. Are they just not charging enough? Are they subsidizing everyone's compute?

  • by Havoc on 4/23/23, 9:55 PM

    >what official affiliation there was between Stability and EleutherAI, they said: “None.”

    That's wild. Handing out a bunch of cash for nothing except vague association. No control? No equity? No IP?

    Reminds me of marketing. Cola puts logos on t-shirts of players and proclaims to have sponsored them, but I have yet to see cola claim to have "co-scored" a goal.

  • by babl-yc on 4/24/23, 3:37 PM

    Why did this post end up getting flagged? https://news.ycombinator.com/newsguidelines.html

    I understand having critiques about the post, but stories about major AI companies fundraising seems on-topic.

  • by EGreg on 4/23/23, 9:23 PM

    Someone help me understand this… the code is open source, so does this mean it’s actually proprietary and we can’t use it?

    And if it is indeed open source with a completely permissive license, then what material difference does it make who originally developed it?

  • by faeriechangling on 4/23/23, 9:42 PM

    Seems like an ungrounded hit piece and it mostly raises questions about who is backing sifted.eu

  • by m3kw9 on 4/23/23, 8:57 PM

    Doesn’t look like fraud but likely lower the valuation unless they can overcome the ip issue

  • by PLenz on 4/23/23, 11:21 PM

    Does anyone have a link to the deck?