[cdcox]

This was probably an okay idea terribly implemented. GenAI creators on social media kind of sense.

Neurosama, an AI streamer, is massively popular.

Silllytavern which lets people make and chat with characters or tell stories with LLMs feeds Openrouter 20 million messages a day, which is a fraction of it's totally usage. Anecdotally I've have non tech friends learn how to install Git and work an API to get this one working.

There are unfortunately tons of secretly AI made influencers on Instagram.

When Meta started these profiles in 2023 it was less clear how the technologies were going to be used and most were just celeb licensed.

I think a few things went wrong. The biggest is GenAI has the highest value in narrowcast and the lowest value in broadcast. GenAI can do very specific and creative things for an individual but when spread to everyone or used with generic prompts it start averaging and becomes boring. It's like Google showing its top searches: it's always going to just be for webpages. Making an GenAI profile isn't fun because these AIs don't really do interesting things on their own. I chatted with these they had very little memory and almost no willingness to do interesting things.

Second, mega corps are, for better or worse, too risk averse to make these any fun. GenAI is most wild and interesting when it can run on its own or do unhinged things. There are several people on Twitter who have ongoing LLM chat rooms that get extremely weird and fascinating but in a way a tech company would never allow. Silllytavern is most interesting/human when the LLM takes things off the rails and challenges or threatens the user. One of the biggest news stories of 2023 was an LLM telling a journalist it loved him. But Meta was never going to make a GenAI that would do self-looping art or have interesting conversations. These LLMs probably are guardrailed into the ground and probably also have watcher models on them. You can almost feel that safeness and lack of risk taking in the boringness of the profiles if you look up the ones they set up in 2023. Football person, comedy person, fashion person, all geared to advice and stuff safe and boring.

I suspect these things had almost zero engagement and they had shuttered most of them. I wonder what Meta was planning with the new ones they were going to roll out.

[dashundchen]

Meta's platforms are already filled with AI slop content farms that drive clicks and engagement for them.

I have a FB account for marketplace, and unsubscribed from all my pages and friends. If I log in, my feed is a neverending stream of suggested rage bait, low quality AI photos, nonsensical LLM "tips" on gardening and housekeeping.

The posts seem to attract tens of thousands of reactions and comments from seemingly real people.

[SoftTalker]

"seemingly" being the operative word. They are mostly fakes.

[Macha]

> Neurosama, an AI streamer, is massively popular.

I think some level of Neuro's popularity is due to the Vedal + Neuro double act though, and some of the scripted/prepared replies.

[cdcox]

Absolutely, and in picking collabs with people who are willing to work with the weirdness and make it funny. Vedal is definitely a fantastic creator to make it work so well and the amount of fine-tuning and tweaking he must do must be unreal. But I think it still shows there is some hunger for this type of content, though you are probably correct that it still needs to be curated, gardened, worked with, and sometimes faked.

[nitwit005]

Yes, it's a success, but not a scalable one.

There probably aren't a lot of decent streamers who are also AI developers around.

[joe_the_user]

This was probably an okay idea terribly implemented.

No, I'd vote terrible idea terribly implemented so good (that it failed).

The argument for GenAI chatbots in culture has to be more than "people like it".

The worst possible GenAI is one that manages to be "better" than the standard sterile, moronic homogenized celebrity that everyone already likes. And sure, like any computer program, a GenAI can be randomly "interesting" but this kind of thing is quite shallow imo.

[gazchop]

> Neurosama, an AI streamer, is massively popular.

This only proves that there's enough people on the planet around that bit of the bell curve to develop an audience.

[fardo]

Aspersions aside, the content’s actually typically pretty involved and has a lot to speak for itself, it’s not low-effort content that one would typically associate with AI.

Laid bare, it’s generally a variety comedy show of a human host and AI riffing off each other, the AI and the chat arguing and counter-roasting each other with human mediation to either double down or steer discussions or roasts in more interesting directions, a platform for guest interviews and collaborations with other streamers, and a showcase of AI bots which were coded up by the stream’s creator to play a surprising variety of games. There’s a lot to like, and you don’t need to be on “that bit of the bell curve” to enjoy a skilled entertainer putting new tools to enjoyable use.

[wussboy]

Is "engagement" the primary metric of modern life?

[wsintra2022]

not the primary metric of modern life but I would agree it’s the primary metric of modern consumer facing business

[CatWChainsaw]

According to your economic masters, yes, because engagement is a proxy for revenue.

[int_19h]

It's fine when AI has its own social media sandbox to play in for people to watch. Stuff like e.g. https://chirper.ai.

I don't think there should be bots like this on social media for humans, though.

[Xplune13]

> This was probably an okay idea terribly implemented. GenAI creators on social media kind of sense.

It boggles my mind that there are people who think this is a good/ok idea. From a human perspective, all it does is pulls the mind ever closer to fictional imaginative world rather than encouraging real life interactions which I believe is inherently wrong no matter what business strategy is wrapped around it.