Models in AI-for-animals

/*
Cost-effectiveness model: animal-inclusive AI work vs. AI pause advocacy.
Originally written by Claude Opus 5.5. Port of animal_ai_vs_pause.py.

Value unit: 1 = an aligned future that is good for all sentient beings.
Doom (misaligned ASI, or another post-alignment catastrophe) = 0.
An OK future that is bad for animals = -badness.
Results are in basis points of that unit per marginal $1M.

Pause advocacy uses log returns: if its annual spending doubled, P(pause)

import "hub:ozziegooen/sTest" as sTest
import "hub:AI-safety/p-solve-alignment" as generalAlignmentCosts

/* Cost-Effectiveness Analysis for ASI Alignment to Animal Welfare */

/* 
Simplified version of https://squigglehub.org/models/AI-for-animals/alignment-to-animals-EV

Unlike the more complicated version, this model does not consider spending-so-far on alignment, it just considers total cost; and the probability of solving alignment is a user input instead of being a derived parameter.

import "hub:ozziegooen/sTest" as sTest
import "hub:AI-safety/p-solve-alignment" as generalAlignmentCosts

/* Cost-Effectiveness Analysis for ASI Alignment to Animal Welfare */

/* In favor of alignment-to-animals:

1. scale of animal welfare is much larger (that's a controversial vale judgment)
2. cost to solve alignment-to-animals is possibly much lower, and unlikely to be higher
3. greater field-building effect of research due to the field being newer