I’m absolutely certain I formulated the solution to the alignment problem.

And I didn’t come to the conclusion I was right by myself, I convinced a CS PHD dude, two persons writing thesis on AI Safety. EDIT: also, literally all people I know.

I also totally convinced GPT4.5 I was right. And I’m good at prompting (like, probably not hugely good, but I know what Dunning-Kruger is, and people have a tendency to see a validation of their genius, interactions with LLMs they don’t really understand).

I want people to believe me, but I wouldn’t, so all I tried to ask to people all along was “Let me f**ing explain”.

People confuse my point and my person (Autism/ADD/DID). I don’t want people to believe me out of a need of validation. I want people to give benefits of doubts enough to listen to my reasoning.

In fact, I don’t like human interactions that much, I don’t want to be famous or even recognized as anything.

Ideally, I’d like to explain thoroughly to someone able to understand and do something with the information.

I can explain with an arbitrary details level. People just won’t let me.

But I’m stuck at “let me explain”. Strangely “please hear me out” sometimes don’t work.

Also, I’m bad at communication, but one thing I don’t get about humanity, is why people formulate counter arguments before being sure that the information has been conveyed.

I mean, when someone explains something to me,I listen, I reformulate to make sure I understood.

People counter on what they thought you say, and end up disagreeing with something I never said.

Also, (perhaps mostly) people make it about my person, my mental health, that’s irrelevant (Though something very relevant on my mental health: I have psychiatric stuff. All diagnosed. So I’m pretty confident I’m not delusional (and my two imaginary friends totally agree)