A human accountability system that records the final human decision before automation begins, preserving intent, assumptions and responsibility.

Remembering the final human decision before automation begins.

LAST HUMAN INPUT is a human-accountability system designed to record, preserve, and display the final human decision made before a process becomes automated or AI-driven.

As automation expands, responsibility often fades. Systems make decisions, but humans originally chose to deploy those systems. This project ensures that the intent, assumptions and acknowledged risks behind those decisions are never lost.

Automated systems influence hiring, healthcare, finance, governance and education.
When failures occur, responsibility becomes unclear.

Most systems record outcomes.
Few record the human decision that started everything.

LAST HUMAN INPUT creates a permanent record of:

• who made the decision
• what they intended
• what they assumed
• what risks they acknowledged

It preserves the origin of automation.

• Decision logging interface
• AI-generated intent summary
• Assumption and risk extraction
• Timeline visualization
• Decision archive
• Anonymous or named submissions
• Permanent record concept

1. A human logs a decision before automation is deployed.
2. The system stores the raw decision text.
3. Gemini AI analyzes the decision and generates a neutral summary.
4. The system displays:
   • Original human input
   • AI interpretation
   • Timeline of responsibility

• HTML
• CSS
• JavaScript
• PHP
• MySQL
• Gemini AI API

Client → PHP Backend → MySQL Database → Gemini AI → Stored Summary → Displayed Record

• Approving automated hiring screening
• Deploying AI medical triage
• Automating loan approvals
• Implementing surveillance AI
• Using AI grading systems

Automation is advancing faster than accountability frameworks.

LAST HUMAN INPUT ensures that when systems act, we can still trace:

who decided, why they decided and what they believed would happen.

This system could evolve into:

• AI governance infrastructure
• Ethical audit trails
• Policy compliance systems
• Public accountability archives

1. Clone repository
2. Import database.sql into MySQL
3. Configure database credentials in PHP config file
4. Add Gemini API key
5. Run on local server (XAMPP / WAMP / LAMP)

Table: decisions

Fields:

• id
• title
• domain
• automation_description
• goal
• assumptions
• risks
• human_name
• is_anonymous
• ai_summary
• created_at

Built for ORIGIN Hackathon — a challenge focused on meaningful technology, human impact and thoughtful innovation.

This project is not about automation.
It is about remembering who chose automation.

MIT License