Models

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

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

invCdf95 = 1.644854

// Helper to extract lognormal params from a 90% CI (a to b)
// σ = (ln(b) - ln(a)) / (2 * invCdf95)
lognormalParams(a, b) = {
  mu = (log(a) + log(b)) / 2
  sigma = (log(b) - log(a)) / (2 * invCdf95)
  { mu, sigma }
}

/*
Expected value of influencing USAID dollars via campaigning for the Presidential candidate who's more pro-USAID, compared to donating to GiveWell top charities.
*/

// actual budget 2025 was $34B pre-cuts
usaid_dollars = normal(34B, 10B)

num_voters = 150M

// relative to GiveWell

// How many rocks on earth?

sigmoid(x) = 1/(1+exp(-x))
logit(x) = log(x/(1-x))

demoSigmoid = {|t| sigmoid(t)}

/*
Comparing the Cost-Effectiveness of generic SysInn FOs situated in EU and US
*/

// Ignoring all variables that do not discriminate between EU and US

// EU COST-EFFECTIVENESS

philtracteu = 0.25 to 0.5 // 1/centralization

/*
Describe your code here
*/

a = normal(2, 5)

//Input parameters
prob_can_buy_galaxies_point_estimate = 3%
fraction_global_wealth_on_galaxies_low = 0.02%
fraction_global_wealth_on_galaxies_middle = 16%
fraction_global_wealth_on_galaxies_high = 100%
fraction_global_wealth_on_galaxies_loguniform = truncate(10 ^ uniform(-3.7, 0),0,1) // Attempt at log-uniform distribution. See right pane for descriptive stats.
fraction_global_wealth_on_galaxies_lognormal = truncate(0.001 to 1, 0, 1) //Log-normal distribution, truncated at 1. Increased lower bound to increase mean/median after truncation. See right pane for descriptive stats.

global_wealth = 878T
donor_wealth = 1B

/*
Describe your code here
*/

// assumption: revenue = api revenue + revenue from enterprise seats + consumer revenue



//////// enterprise ////////
// key uncertainties: 

// assumption: revenue = api revenue + revenue from enterprise seats + consumer revenue



//////// enterprise ////////
// key uncertainties: 
// - what is the pricing for these 8M enterprise seats?
// - any other kinds of enterprise deals that this does not account for

num_seats = 8M // Q4 2025 earnings

/*
Generated by Squiggle AI. Workflow ID: e3c09385-d27b-4836-95d9-d0afa5684643
*/
// Cost-Effectiveness Analysis: ITN Distribution for Malaria Reduction in Kenya
// This model analyzes the cost-effectiveness of distributing long-lasting insecticide-treated nets (LLINs)

import "hub:ozziegooen/sTest" as sTest

// == Core Epidemiological and Demographic Inputs ==

/*
Describe your code here
*/

/** Wu & Khlangwiset baseline cost per DALY ($/DALY) */
wk_baseline_cpdaly = uniform(0.21, 2.08)

/** Real-world adoption rate
    Point estimate: 7.2% repeat use in Malawi (Nakoma Ngoma et al. 2025)
    Uncertainty: plausibly 5-15% */

/*
Capturing countervailing dynamics that affect the relative tractabilities of Innovation advocacy in US vs EU
*/

// EU relative strengths
stability = 1.3 to 2
neglectedness = 1.125 to 2
funding_additionality = 1.125 to 2
policy_surface = truncate(normal({p5: 2, p95: 6}), 0.6, 7.4)
activity_additionality =  1.3333 to 2.0833 // AKA grantee influence

/*
Generated by Squiggle AI. Workflow ID: ca63c333-eaa6-40be-a4b5-798a2ba5e8d4
David Reinstein -- edited from this starting point. The original model started with https://acbmcostcalculator.ucdavis.edu/ (Risner et al 2021)
*/
/// Initial code generated by GPT5 based on conversation at https://chatgpt.com/share/68b9b37b-810c-8002-9300-1f6c6a8da252 on 4 Sep 2025.
/// [NB: that link stopped working -- https://chatgpt.com/c/68e39548-8240-8011-b6d0-865fb359bd23 is the internal link that moves that ACBM model

/// Cultured Meat Cost (CM-COGS) — Simple Scenario Model (v0.1)
/// Audience: economists & non-engineers
/// Epistemic status: starter scaffold with reasonable ranges; replace with your own data.

/*
Describe your code here
*/

a = normal(2, 5)

/*
Describe your code here
*/

// 1. SCALE PARAMETERS
chefs_trained = 50
meals_per_chef = 5000
impact_years = 10
total_meals_per_year = chefs_trained * meals_per_chef

import "hub:ozziegooen/sTest" as sTest

// Model: Cost-Effectiveness of Institutional Meat Reduction

@name("Inputs")
inputs = {
  @name("Meals Replaced in 2024")
  @doc("Total number of meals replaced across all institutions in 2024")
  mealsReplaced2024 = 2474256

import "hub:ozziegooen/sTest" as sTest

// Model for cost-effectiveness of cage-free corporate campaign
@name("Key Inputs")
inputs = {
  @name("NPV of Hens Affected by All Commitments")
  @doc(
    "Total number of hens that would be affected if all corporate commitments are fulfilled"
  )
  npvHensAffected = 90932653

// ==========================================
// FUNDING SCENARIOS
// ==========================================

// --- SHARED PARAMETERS ---
baseSalaryY1 = 175k
baseSalaryY2 = 182k
benefitsBurden = 0.4
fullyLoadedY1 = baseSalaryY1 * (1 + benefitsBurden)
fullyLoadedY2 = baseSalaryY2 * (1 + benefitsBurden)

/*
Describe your code here
*/

a = normal(2, 5)