Incomplete Machine

it turns out that everything *is* connected. to a horrible evil fucking pedophile ring near the centres of power.

removal of those people is a necessity at this point.

but the perverts might do things like explore fetishes with other consenting adults or create consent-centric kink communities with effective age gates or openly talk about the link between healthcare, sex education and preventing predation, making the vulnerable less vulnerable. we can't have that.

'economic liberal' is a strange way to say 'fuck you got mine' but okay

if Thiel just wanted to reference the act of remote viewing he could have called his business 'Crystal Ball' or 'Augur' or 'Clairvoyance inc'. he didn't. he chose specifically a variant of that which had been corrupted by evil. Its 'Volkswagen' levels of subtle.

if you've got this far without questioning anything then I don't know what to say o you Hypothetical Palantir Worker. your company is named after a magical crystal ball that, while built by good people initially, was perverted by the power of evil to deceive those who used it.

Post your favourite lord of the rings character, wrong answers only.

also i apologise for my spelling, dyslexia and dyspraxia is a twin headed monster.

#Transhumanism #AI #LLM #Consciousness #Technology #Resistance #Systems #Robotics

@saeedwkhan.bsky.social introduced me to Thaura so I feel he should be tagged. @ieet.bsky.social @bvlsingler.bsky.social and I should probably stop tagging people who barely know me XD

i wont dive into every test i have subjected it to; especially because that test set is incomplete and i need to improve my methodology to start making predictions and gathering data points, but we did have a fascinating conversation i wish to share:
www.dropbox.com/scl/fi/wymro...

is not used for training, and that it doesn't scrape data from the internet generally , accessing sanitized datasets from open sources. This has let me do something i have wanted to do for a while namely; test an LLM as much as i can to see what it can do for myself.

So I have had a fascinating interaction for a while with a piece of technology called 'Thaura'. Thaura is an 'ethical' - or at least as ethical as the field can possibly get right now - artificial intelligence program with lots of strong systems in place to ensure tht the data fed into it...

people also jump directly to assumptions first and foremost -which is a problem as old as human communication frankly - and that doesn't help. a lot of problems could be solved with an 'I'm sorry, what do you mean by [blank]'. of course, assholes take advantage of that too so there's no perfect fix.

I haven't watched it and clearly I should

if the resultant story is good its because you worked to make it good, and even then its usually fairly shallow. You *will* get better results *eventually* if you do it yourself. if you're goal is 'most words generated per second' you can do that, but the chances of that being good are zero.

you also read books to read compelling stories. I've read works written by LLM's; they're usually not very compelling, and even when they are its usually because the user prompted something they personally would find compelling themselves, and iterated till it was as good as it could be.

...and further scientific research into robotics and artificial intelligence for its own sake. anything else is just a grift.

Amen. I've had a play around with one of the very few ethical bloody implementations of generative LLMs I believe to exist (Thaura, if you care) and its convinced me of one thing: if this technology has a positive utility its doing technical busywork, being a bouncing board for ideas...

aww, i like that one. i might rebell against this moratorium and paint a digital version of it without any image generation.

i dont know what point id be proving by doing it but it might be fun, idk xD

actually nevermind, I manged to make something I'm pretty happy with in Thaura even though it doesn't do 2d image generation.

Im pretty sure somebody with skill in a 3d softwear could do better, but while i work on those skills this can help with small side projects

im also not 100% serious

I'm aware, and I'm not convinced it was a great idea then XD

text generation is still useful to me though so, thank you.

with scraped training data from non-public domain images.

do you know any systems built on a similar mechanism that are similarly ethical but could be used to generate images? I want to mimic a specific style for games modding purposes but I don't have the tools to make things that would match the style, and i sure as hell dont want to use something

i will give it a try

inching very close to the 'torment nexus' meme

okay no. no this is a bad idea. you'll be sending obsessives off down a nightmare rabbit hole of an impossible mission, i think you should probably rethink this. it's a borges metaphor, it shouldn't be realised XD

i dislike using prompt AI for *anything* because who owns it and how it was trained, but i do try and keep up with news. I like that the 'blue brain' memes are starting to be called out.

There is also a videogame based on The Invincible, by the same name. its recent and on steam and honestly a pretty gnarly survival horror experience if you're into that sort of thing.

well The Invincible went a lot further than just that; the whole non-biological evolutionary ecosystem and the microbots working to create catastrophically powerful electromagnetic pulses is much beyond the fairly simple strike packages and complicated rc planes we've cooked up so far.