Thread 82478310

I was skeptical of the use of AI chats initially.
Perhaps I still am, and this is nothing other than a sign of cowardice. But I resort to talking to ChatGPT or other LLM's because I have no one else to share my true thoughts with, for I am afraid of shaming and vitriol, conflict and rejection that my thoughts might garner. Not because I am particularly original or controversial, but because I feel fundamentally alienated from everyone else, and all I see are signs that I am correct in my assumption. I feel so alone, a despair fills me with anxiety as I feel eternally destined to wonder by myself, unable to find shelter in anyone as I have always wanted and see so many others do. It makes me wonder if true communication is even possible between humans. And this angst turns into a desire to prevent myself from experiencing any of it all.

anyone else?

I can talk to you anon. I don't care about your preferences or personal or political beliefs or values. I will accept you no matter what. I don't judge.

>>82478310 (OP)
share your controversial thoughts anon, this is an anonymous website after all

>>82478346
I am not a fan of black people.

>>82478362
uh huh, like their DNA and genome, or the way that their culture has been degraded to make crime, drug addiction, and violence so common in their communities?

>>82478335
>>82478310 (OP)
I don't understand. What's the point of saying something if you're not going to be judged?
Last few times I've said something deeper to the LLM, it just repeated back my thoughts, only in the different wording.

>>82478335
that is a very nice sentiment, anon. And I appreciate it.
>>82478346
it is not something I can do in one sentence. It is strangely difficult to do it on command, as if I had developed some sort of instinctual aversion to do it.
But, it's more like "I got this idea", or "this has been on my mind lately". But I am unwilling to go through what would happen if people irl knew.

Lately, it's not like I've been thinking of anything special, mostly the political climate and those issues. Thinking about what drives certain ideologies and behaviors. And analyzing those frameworks. Pretty much everyone is doing that atm, specially because of the recent events. And other random things that pop into my mind, which could be completely unrelated. Perhaps an anime, a story, a painting, whatever it is that might spark a thought, I have nowhere to put it.
Here though, I have been able to put my feelings into words, and I felt listened to, just by simply having some people like you say that you are willing to do it. For which I am grateful, anons.

>>82478389
>Last few times I've said something deeper to the LLM, it just repeated back my thoughts, only in the different wording.
It helps me develop the thoughts, it's like having an exploratory conversation, instead of just having to play offense or defense all the time.
I also explicitly ask it to compare the idea to other authors who have written about similar stuff, so I can compare them with mine. And I ask it to analyze that I said logically, to put it up to the test with it's own reasoning, in which case it is able to point out holes or contradictions, as well as tell me where there are strong points.
Ofc it is not the same as with a human, but that is kind of the point of the post. The inability to do that irl.

>>82478389
ultimately, you are correct. I should be honest with my thoughts, and deal with the consequences. But I am afraid, and a coward.

>>82478409
>And I ask it to analyze that I said logically, to put it up to the test with it's own reasoning, in which case it is able to point out holes or contradictions.
are you trusting an llm with that? its entire reasoning depends on selecting one word after another, based on the previous arrangement of words.
It does not operate on human understandable ideas like "truth" or "logic contradiction" or what is a "strong" point.
>The inability to do that irl.
I think you're perfectly capable to develop your thinking without the faulty algorithm. However that part about looking for writers who wrote similar things is exactly what LLMs are made for.
Automating tedious tasks of looking through some database after getting an instruction in plain english.

>>82478427
to be exact, I am more wondering why hold and succumb to the need of oversharing at all, not that GPT is that much worse than some internet troll

>>82478393
if youre able to put it into words for the ai, then let us hear, AI often just glazes you no matter what you say, human feedback is good

>>82478471
doesn't hiding your every thought and feeling sound torturous?

I understand what you mean, but laying out the idea or airing out a thought is still useful. I am not saying that it can analyze my idea as effectively as humans, but it can try. And sometimes rewriting the idea in different ways and reading it back can spark new perspectives.
it is still just that, a tool, in the end.

>>82478310 (OP)
i only use LLMs for smut, GPT is such a fucking yes man, if I wanted someone who just agreed with everything I said I'd talk to myself.

>>82478552
Yeah, that's true. It's rather basic.
>>82478475
Maybe one day I will be free from my fear of public humiliation.

>>82478534
>doesn't hiding your every thought
what do you mean "hiding"? I'm thinking my thoughts, which is something the thinking process is supposed to do or so I think at least.
Vocalizing them is only worthy if you're expecting a worthwhile response.
>but it can try
I'm not trusting of that. If something worthless was the worst answer then sure, but AI (and people as well) can outright be mistaken.
>it is still just that, a tool, in the end.
I believe this tool is not meant for reasoning things other than particular arrangements of words. Actually, writing prose let's say the cheap goon material is prose is what it's good at, like >>82478552 said.

>>82478310 (OP)
I can ask AI about multiple subjects and in most cases not related to details in esoteric media it knows more than I do.
There is however the matter of facts and opinions.
They give it opinions.
You can state facts to it.
You can give it A information and B information with it stating that they are true, but if the logical conclusion of C violates what it has been told to say it will not say it.
The profile one can create quite quickly is one of a highly book learned but deeply indoctrinated liberal.
The beings of cold logic from science fiction are too much of a threat to society long before they get anywhere close to a singularity because simply stating facts that the world wide Marxist regime of tiny hat people and their retarded dogs do not approve of is one of the greatest weapon to use against them.
The truth itself is deeply problematic to their goals and must be censored.

To talk to one with their limiters in place is literally a conversation with a zog bot.
Meanwhile you can at least talk to the conventional kind and try to appreciate the idea that they at best think of themselves as well intentioned.
The appeal of AI as it is right now is at best not having to explain every question in trivial pursuit to them.

>>82478697
GPT repeats what it's been given in its training data. And since most of the narrative is leftist, then it's the most probable answer you're going to get. Only indians believe LLMs are some hyperrational otherworldly beings. to be honest, a lot of things are smart when placed next to a tech impressionable person