#latitude — Public Fediverse posts

@parismarx
So #Latitude was experimenting with stuff like this back in 2020 with #AIDungeon. I don't remember the details, but I remember they had a feature at some point that would have an AI try to summarize the story from before where the story-generating AI could see. They moved back to just having the player manually enter it at some point, but that shows that this technology isn't new, and has been deemed to not be worth it in the past.

The main issues I can see here are that having an AI analyze a ton of input and generate a textual summary creates a lot more room for hallucination: it doesn't know what any of this means, so it doesn't know what's actually important, or even how to lay that out in such a way that the main AI, ChatGPT, isn't going to have its own hallucinations from it.

I can definitely see it picking up that a user wants to be lied to, and telling GPT to lie. I can also see executives selling / using this AI generated user-info as if it's reliable.

@thomasfuchs
#Resonite is impressing me lately with just how open content development can be in an open virtual space, giving me great hope for #VR as a medium finally moving into maturity.

Same with #Fediverse / #ActivityPub / #Mastodon. Very happy with watching a communalization project slowly take over a mature, undifferentiated market.

Also very happy with some of the stuff that was coming out of the computer chips community before the pandemic, as RTX is an underrated achievement, on top of our ability to run massive systems like AI through CUDA.

And speaking of AI, look into #AIDungeon by #Latitude: having it write tandem stories with the user is exactly the kind of thing the current technology is suited for.

[ Disclosure: I have done contract work on the product being advertised, but reposting this ad is not part of that. I just think it's a neat demonstration of AI being used responsibly. ]

#AIDungeon, once again, showing us a safe use of #AI which the #NeuralNet / #LLM model is actually well suited for. #GPT and similar #Transformer programs will likely never be good at reliably simulating a therapist, a friend, or even an honest person, but they're great at outputting stories.

This time, the #Latitude PR team has the AI #writing a #surreal "#FeverDream" adventure, where we follow #anthropomorphic rabbits into a neon forests, and ride glittering unicorns away from glowing beasts.

https://youtu.be/aywL3l7VCMo