Skip to content

Instantly share code, notes, and snippets.

@automenta
automenta / compsynth.js
Created April 3, 2025 14:51
Comparative Synthesizer
/*
Comparative Synthesizer
To use this script, follow the prompts:
- Enter your initial prompt
- For each variation:
- Copy the prompt
- Paste into browser
- Paste result back into terminal
- Answer y/n for more variations
import com.fasterxml.jackson.databind.ObjectMapper;
import jdk.jshell.JShell;
import jdk.jshell.SnippetEvent;
import jdk.jshell.VarSnippet;
import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;
import java.util.*;

S list questions that one might be afraid to learn the answer.

Here are some examples of questions that one might be afraid to learn the answers to:

Do I have a terminal illness or serious medical condition? Has my partner been unfaithful? Will I lose my job or get fired soon? Am I going to go bankrupt or face financial ruin? Did a loved one die in an accident or tragedy?

https://www.dhs.gov/medialibrary/assets/videos/23065 AUDREY - Public Safety Agencies Pilot Artificial Intelligence to Aid in First Response | Homeland Security Skip to main content Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS

https://dataverse.jpl.nasa.gov/dataset.xhtml?persistentId=hdl:2014/53408 Trusted and exPlainable Artificial Intelligence for Saving Lives (TruePAL) Technology for First Responder Safety Project Introduction - Root No abstract provided but the video is more like advertisement to push the project :)

https://www.spiedigitallibrary.org/conference-proceedings-of-spie/12101/121010G/Neurosymbolic-hybrid-approach-to-driver-collision-warning/10.1117/12.2620209.pdf Neurosymbolic hybrid approach to driver collision warning - SPIE Digital Library

saveModel: False
system:
seed: 0
data:
block_size: 1024
model:
model_type: gopher-44m
activation: LeakyReLU(negative_slope=0.01)
n_layer: None
n_head: None
saveModel: False
system:
seed: 0
work_dir: ./out/chargpt
data:
block_size: 1024
model:
model_type: gopher-44m
activation: LeakyReLU(negative_slope=0.01)
n_layer: None
saveModel: False
system:
seed: 0
work_dir: ./out/chargpt
data:
block_size: 1024
model:
model_type: gpt-mini
saveModel: False
system:
seed: 0
work_dir: ./out/chargpt
data:
block_size: 512
model:
model_type: gpt-mini
saveModel: False
system:
seed: 0
work_dir: ./out/chargpt
data:
block_size: 512
model:
model_type: gpt-mini
saveModel: False
system:
seed: 0
work_dir: ./out/chargpt
data:
block_size: 512
model:
model_type: gpt-mini