I worked on this intimately, the level this blog post misses the mark is amusing. It's not my place to go into it, so I'm afraid you'll have to take it on good faith, that this blog post really isn't even what happened, never mind not even really how the tech worked etc. Skylar and Brian who build it are brilliant, and this blog post is super shallow, at least in it's facts about how it worked and why it ended. I don't know it invalidates the post, only that, well, it's not really factually accurate, the guessing is almost flat wrong in parts.
Just want to say thank you for your contribution. A lot of the wonky stuff made it really funny to watch, but at the same time impressive (for what it was).
Do you have any links to share about details surrounding the project? Would be fun to understand more about the process but as well the work in general.
A lot of it is not mine to share, I was basically a CXO hire brought in by Skyler Hartle and Brian Habersberger who created it* to work with them on assessing the viability of the product they'd built behind the scenes, as a business, and it's levels of scale etc. Skyler and Brian are amazing genius level dudes (and also HN users!). I know the whole story end to end, pretty much exactly as it unfolded, it's not really anything like this blog post thinks it is, to the point I don't think I even agree with the author. Again: I'm sorry but much of it is not mine to share, for a great many reasons.
*As a side project, over years, and years, before it took off, "over night", there is a lot of tech unmentioned they built totally unrelated to LLMs etc, much of it still worth, not mentioning, because it's "cutting edge" still even today)
> Anyone affiliated with the project can email me privately and I'll add notes where appropriate.
This is the principal difference between traditional journalism and modern blogging.
In a real newspaper, a blog post like this would have been followed by "Mismatch media did not respond to a request for comment", where the request for comment was made in good faith (AKA not 30 seconds before hitting "publish. All claims would have been properly sourced and have included annotations like "according to three former employees of Mismatch, who spoke to us on the condition of anonymity as they're contractually forbidden from discussing the project".
There's a lot of good that modern blogging can do in the world, the fact that anybody can publish a blog means that important stories aren't subject to gatekeeping, but it's a real shame that this transition means losing the journalistic standards that we historically have had.
My dude, you're the one who wrote this post (presumably) so you're the one who should be reaching out to the relevant parties to seek out information, not the other way around...
Neurosama is a very popular vtuber and runs on Twitch without any particular controversy about the fact that she's an AI, I believe an artist even made her model for free.
I think she even became the most popular "female" streamer on the Twitch platform for a while.
I admit I wasn't aware of that channel when writing this post.
That said, VTubing is a medium predicated with several layers of irony (I also admit I don't fully understand non-AI VTubers) and the implicit assumption among viewers that it's inherently fantasy, so the uncanny valley logic and mainstream appeal that this post frames is not as applicable.
>so the uncanny valley logic and mainstream appeal that this post frames is not as applicable.
To be the most popular female streamer on Twitch and the most popular vtuber on any streaming platform for an entire week means you're already pretty mainstream, if not, there's really nothing mainstream on Twitch.
There was something really amusing with how oddly broken and nonsequitur the show was. It was the Seinfeld you loved but the jokes were bizarre, dadaist, and would end awkwardly in the middle of a scene after an uncomfortable pause. It was addictive in how unpredictable it could be (well, when jokes weren't repeating). And you're right - the energy was fantastic. I loved twitch chat spamming memes like NEW <x>!!! Or going insane every time Kekler showed up. Or spamming MMMM when the extremely loud microwave turned on.
I talked about Neurosama in another comment but there is also DougDoug, a really popular streamer and content creator that very often use AI chatbots to make content.
Always without any particular controversy about the AI chatbot.
Many of the most popular streamer also use AI voices of famous people to read donations.
And you created a wonderful cautionary tail in working backwards when you're smart but wrong. I'd imagine that post took a little while to write, and given how little you know about what you where talking about, is why I said... not to be mean, just... it really does become amusing.
> Would a 24/7 AI-Generated Twitch Stream Even Work Anymore?
Yes. There is right now a 24/7 live discussion happening right now around posts on Twitter / X with a panel of.....AI characters giving their expert opinion on a Twitch live stream. [0]