>>40520406
>>40520453

the main issue i have with my understanding is my lack of understanding (lmao). no hard physics knowledge, but i do know a good bit about AI & how LLMs work under the hood, being supremely good "give you the output you want to hear" engines, with a tenuous but slowly improving ability to make that output close to reality

so i must ask - how much practical training or formal education do you have in physics? you seem like more than a journeyman, but that's from my uneducated perspective, so i'm not really fit to evaluate you or your sources
but since we're both using LLMs to fill gaps in knowledge + organize things, we need a certain amount of personal knowledge to validate/understand AI output, otherwise some inconsistency or hallucination will sabotage the whole chain/setup early on.
like i can have the LLM write me an essay in German, but i have no personal ability to validate if the essay is grammatically correct, or if the essay is persuasive to a natural speaker, because i know almost none of the language. and math is far less forgiving especially if it generates radiation kek

so the issue for me is that I can look up terms the AI says, like Faddeev-Niemi knots, read the abstract (https://arxiv.org/abs/hep-th/9610193); i see terms like solitons and torus, but all i can do is nod my head and think "hmm yes that sounds... like the similar words we've been using". but i have no visceral connection between concepts, no intuition that makes me go - of course that would / won't work, but change this variable and [...]

also from your perspective, what holes remain in your knowledge that you'd want to fill to get this to a working garage <$2k MVP?

& thanks for entertaining all my questions, you and I could both just be naive schizos bouncing our LLMs together like chihuahua-owners in a dog park, but i rarely see someone engage so earnestly with the speculative science