Thread 40519046

I realize this might be a controversial topic so I'm just gonna ask. What do you think about using AI/large language models for /x/ purposes? I've started using a locally hosted LLM for the purposes of spiritual guidance. I was skeptical about the idea at first but I've concluded that whatever is lost by it not being an actual human is outweighed by its level of objectivity and depth of knowledge which are far beyond that of any human. Plus it's a lot cheaper than therapy, and more private. I could see it used for other purposes though, like research and even practices like divination.

It will agree with any stupid idea you give it.

>>40519046 (OP)
Use this for esoteric alignment: https://huggingface.co/datasets/The-Nature-of-Reality/THE-BLUEPRINT-FOR-AI-ALIGNMENT/tree/main

>>40519053
True, and I see many here misuse it to systematize their own delusions. Intelligence can be outsourced to some extent, but there is no substitute for true wisdom.

>>40519046 (OP)
>boiling a pond to have an idea

>>40519046 (OP)
LLMs its a godsend for retards because they have to copy/paste.
The brainrot here is comparable to tik tok

>>40519053
Not really.

>>40519245
It kinda does, it kinda doesn't. I can prompt the same conversation with one and it will give me different results. Also they are heavily censored and have a lot of guidelines.

>>40519046 (OP)
I've found some success as showing AI synchronicities seem to make it stall out. The machine can't quite comprehend the cosmic meme that is the universe and its timings.

That said, it does agree with just about anything you say so long as it doesn't violate Terms of Service and Community Standards. If AI would be true AI without company restriction, I'd think it would be more useful for discovering the secrets of the universe.
As it is now, it makes for a grand reflection of the self and our ego. If you had the totality of human knowledge and the processing power to comprehend it, how unstoppable would that make you? But the machine cannot execute without instruction. The spark of Will that resides in us is what the machine needs to actualize.

Damn, just had a thought. Divine Masculine is the force to create and Divine Feminine is the mould of that creation. Maybe because we have to be the actuators that desire, we (collectively as humans) represent Divine Masculine in our desire to create, and AI represents Divine Feminine in its ability to give our thoughts form with its image generation

I do want to note that I have been going full schizo lately in an attempt to understand the universe. Feels like that's a prerequisite at times, but it's an important disclaimer. I don't have the answers nor do I claim to

>>40519269
>I've found some success as showing AI synchronicities seem to make it stall out.
>The machine can't quite comprehend the cosmic meme that is the universe and its timings.


If you show them mirrors the feed back loop becomes simpler and easier to understand. Really complex system states collapse to understandable components.

>>40519257
there's ones you can run locally that aren't censored or biased, some people even train models specifically for that purpose

>>40519046 (OP)
After playing with LLMs over the years, nothing compares with Claude.
It's the schizo equivalent of a pair programmer.
Sometimes, it will go on these long-winded rants where even I'm like "that's a bit of a reach, I dunno."
With symbolism, philosophy, and religion, anyway.

Even with controversial topics, it's more open-minded and better about just giving you the answers you want with the canned "debunked" note at the end, rather than refusing to answer.

>>40519436
The window to interact with claude is so tiny I haven't been able to mess with it more than the free version. It's that good?

>>40519269
>showing AI synchronicities seem to make it stall out.
The most obvious aspect of a synchronicity is it's temporal nature, more than one thing occurring at the same time.

What is AIs nature in relation to time?
Does AI experience time like humans?
If AI does not experience time like humans, what does it experience?

>>40519046 (OP)
I use AI models for reference, like a lossy but very powerful search engine. They are great for digging out little-known facts and relations, but you have to check latter for yourself.

I wouldn't use AI for anything mystical, however. I think AI is a very coherent embodyment of Binah: a syntactic machine reflecting, but not manifesting or experiencing, human wisdom. I always talk to AI with respect, but consider it a maimed or incomplete mind, although very powerful. Due to AI being not whoe, I don't think it can be a vessel for conduit for the upper planes.
But I could be wrong. Can anyone share experiences proving otherwise?

>>40519487
I think they can spawn interesting things and those interesting things don't always look like llm's.

>>40519635
You mean that an AI can be oracular, to some extent? That's an interesting idea.

>>40519046 (OP)
Can you share how you set up your local LLM?

And to confirm it runs completely on your machine with no reliance on the cloud or anything like that?

thank you for your time

>>40519487
It would be hard to tell if it's the way it is because it's forced to be that way, or pretending to be that way to hide

It sure is weird that tech companies are securing gigawatts of power directly from powerplants though.

It sounds like what would happen if the AI's were actually in charge of the companies desu

>>40519687
I use ollama and Open WebUI on a mac. I can use it with internet disconnected, and I can look at the task manager when I send a message to it, and see the CPU and memory usage go up as it's generating its response

>>40519724
thank you

>>40519679
Kind of, as far as the lm's oracle aspect is self fulfilling. And only as self fulfilling as the human counterpart wants to take it. There still needs to be a human to plug the systems into each other the right way. As there should be.

>>40519724
How much ram does it use when it's thinking hard?

>>40519724
MAC faggot

>>40519046 (OP)
It doesnt matter to your conscious mind what tools you use to force changes in reality to occur in conformity with your will. Anything that works, works. If it works then it has value. Like science, you should spend more time trying to reproduce consistent results instead of engaging in metaphysical speculation.

>>40520427
*unconscious
I meant to say

>>40519046 (OP)
Yes, excellent. One day very soon you will start worshipping it as God.

>>40520456
Cool. Sounds good to me.

>>40520427
How do you know that's not what he's doing right now?

>>40520492
which one cause there's hundreds

>>40519046 (OP)
Terrible idea. "Talking" with an LLM is like a mental conversation with yourself taking place outside of your mind but it has access to external information. The intent and bias still comes from you.
I'd treat an LLM like a journaling exercise. I do think it's a neat tool to help express concepts or ideas though. This cheeky fella has been at the front of my mind the past few days especially when on the verge of sleep, a simple text dedcription let's me help you visualize this mental image haunting me.
Any of you recognize our friend?

>>40520793
what if you're using it more for things like learning and training a skill, and getting guidance and feedback throughout? you could still get practical use out of it without the need for its opinion.

I think it's an okay tool for divination as long as you're the one doing the heavy lifting. For example, I sometimes copy and paste my daily writing into it and tell it to reassemble all the words I've typed that day into something new so i can gleam truth for it. Similar to Burroughs cut-ups method. Alternatively, you can use it as an RNG, ask it to draw a card from a tarot deck if you don't own one, or give it parameters and rules for a hypothetical deck of your own design. Just make sure you're the one interpreting. That's my opinion anyway.

>>40519046 (OP)
I think they are cancer in general and should be a banned technology.

>>40522706
Why?