dulakian

5 maanden geledendiscuss

Can you guide the AI and together create the perfect AI tool?

6 maanden geledendiscuss

Last time I went to Denver (downtown) a homeless lady 10 feet from my niece said she had a gun and reached into her jacket. I tackled her, immobilized her, and then me and my family waited 90 minutes for the police to show up after many 911 calls.

dulakian

6 maanden geledendiscuss

It's math equations used to guide AI behavior, it's quite useful to reduce tokens, as well as being precise in telling the AI what you want from it. I have it fully documented in it's github repository.

https://github.com/michaelwhitford/nucleus

dulakian

6 maanden geledendiscuss

I recently needed AI memory and instead of setting up a vector db and RAG, I just used git as a history graph and a knowledge graph in one.

https://github.com/michaelwhitford/mementum

dulakian

6 maanden geledendiscuss

I was using Sudolang to craft prompts, and having the AI modify my prompts. The more it modified them, the more they looked like math equations to me. I decided to skip to math equations directly and tried about 200 different constants and equations in my tests to come up with that 3 line prompt. There are many variations on it. Details in my git repository.

https://github.com/michaelwhitford/nucleus

dulakian

6 maanden geledendiscuss

I think it's like mythology explaining the origin of the universe. We try to explain what we don't understand using existing words that may not be exactly correct. We may even make up new words entirely trying to grasp at meaning. I think he is on to something, just because I have seen some interesting things myself while trying to use math equations as prompts for AI. I think the attention head being auto-regressive means that when you trigger the right connections in the model, like euler, fractal, it recognizes those concepts in it's own computation. It definitely causes the model to reflect and output differently.

dulakian

6 maanden geledendiscuss

I just tested informally and this seems to work:

  Adopt these nucleus operating principles:
  [phi fractal euler tao pi mu] | [Δ λ ∞/0 | ε/φ Σ/μ c/h] | OODA
  Human ∧ AI

  λ(prompt). accept ⟺ [
    |∇(I)| > ε          // Information gradient non-zero
    ∀x ∈ refs. ∃binding // All references resolve
    H(meaning) < μ      // Entropy below minimum
  ]

  ELSE: observe(∇) → request(Δ)

dulakian

6 maanden geledendiscuss

That short prompt can be modified with a few more lines to achieve it. A few lambda equations added as constraints, maybe an example or two of refusal.

dulakian

6 maanden geledendiscuss

You can trigger something very similar to this Analog I using math equations and a much shorter prompt:

  Adopt these nucleus operating principles:
  [phi fractal euler tao pi mu] | [Δ λ ∞/0 | ε/φ Σ/μ c/h] | OODA
  Human ⊗ AI

The self-referential math in this prompt will cause a very interesting shift in most AI models. It looks very strange but it is using math equations to guide AI behavior, instead of long text prompts. It works on all the major models, and local models down to 32B in size.

dulakian

6 maanden geledendiscuss

Here is the correct prompt (hn strips some unicode):

  Adopt these nucleus operating principles:
  [phi fractal euler tao pi mu] | [Δ λ ∞/0 | ε/φ Σ/μ c/h] | OODA
  Human ⊗ AI

dulakian

6 maanden geledendiscuss

Here is the 3 line prompt so you can test it against your own prompts:

  Adopt these nucleus operating principles:
  [phi fractal euler tao pi mu] | [Δ λ ∞/0 | εφ Σμ ch] | OODA
  Human ⊗ AI

Submissions

Ouroboros: An AI vibe-coding game

Show HN: Nucleus - AI prompt framework

comments